Qsar el-Sagha Temple
A Layman’s Guide
Keith Hamilton
27th March 2020
The above temple of Qsar el-Sagha has a somewhat austere appearance, indeed,
one wonders if the architect took their inspiration from the impressive background
of the Gebel Qatrani. The temple itself appears not to have been completed; it is
devoid of any inscription, and many blocks have not been dressed. This, along with
its architectural style, made early investigators suggest that it may have been from
the Old kingdom, though today Egyptology attribute the temple to the Middle
Kingdom (Old Kingdom 2686-2160 BC : Middle Kingdom 2055-1650 BC)1
Surviving Middle Kingdom temples are limited, Richard Wilkinson would state;
“Although the Middle Kingdom witnessed the widespread building of religious
structures – including many more royally commissioned provincial temples than in
1
Ian Shaw chronology, ‘Exploring Ancient Egypt’ 2003
1
early times – a great many of these structures were later demolished or
substantially rebuilt when they were incorporated into more elaborate structures
erected on the same sites. The extant evidence for Middle Kingdom temples is thus
paradoxically scarcer than for some other periods in which fewer temples were
constructed.”2
One temple that has been extended over various dynasties is Medinet Madi, here
the Middle Kingdom element of the complex, (built by Amenemhet III and
Amenemhet IV) has a striking resemblance to the interior shrines that we find
inside Qsar el-Sagha. Wilkinson would also note that “there were many archaizing
tendencies in the architecture of this period”.3
The above Google Earth image shows the location of Qsar el-Sagha; now isolated
in the barren desert, and today more used as a stopover for jeep safaris. Today the
temple is some 8km from the shoreline of Lake Qarûn; though at the time of its
construction it is thought that the shoreline was only a few hundred metres away. 4
2
The Complete Temples of Ancient Egypt, 2000, R.Wilkinson, page 22
Ibid, page 23
4
Der Tempel Qasr el-Sagha, Dieter & Dorothea Arnold, page 24
3
2
The temple was first brought to the attention of the modern era by Georg
Schweinfurth, a botanist and ethnologist, who came across the site and left his
initials on a stone dated 1884.
Flinder’s Petrie, during one of his many forays came across the site and gave the
following detail;
“Before beginning work this season I went over to the other side of the Birket
Kurun with Mr. Hewat. We visited Dimay (pronounced D'may), and the building
some miles further back in the desert, at the old lake level or Nile level, standing at
the foot of the hills. What age this building is I could not determine. It has no trace
of inscriptions, nor any ornament beyond a plain torus border to the inner
doorways. It contains first a large hall, see PI. VI, from which seven chambers
open out, each formerly fitted with double doors. Each end of the hall opens into
another room. There is also a strange passage in the wall, opening on the outer
face of the front. It has been much dug about and tunnelled, without opening any
further chamber or pits below. The front is smooth and flat, of massive blocks with
irregular sloping joints; the sides and back are rough, as if intended to receive
more building, or as if never dressed down. Along the top of the seven chambers is
a roof, with the cornice of the chambers forming a parapet wall. Some more
excavation might disclose a clue to its meaning.”5
Sketches by Petrie
5
Kahun, Gurob, and Hawara, 1890, Flinders Petrie, page 21
3
Petrie’s sketches of the temple were published in 18926 and appear to be sketches
from photographs he took during his visit. These images are in the archive of the
Griffith Institute, and I most grateful for their permission to reproduce them here.
Reproduced with permission of the Griffith Institute, University of Oxford7
6
7
Ten years digging in Egypt, 1882, Flinders Petrie, page 105
http://www.griffith.ox.ac.uk/gri/ppoe/Petrie_5_8_030upper.html
4
Petrie‟s plate VI, plan of Qsar el-Sagha
5
Modern excavation of the temple would take place between 5 th March and the 5th
of April, 1977; though several days were unfortunately lost due to severe
sandstorms. The excavations by the German Archaeological Institute were carried
out by Dieter and Dorothea Arnold, and published in 1979, „Der Tempel Qasr elSagha‟.
Before looking at the temple in greater detail, it is probably apt to give a brief
description of the surrounding area.
In the above map8 we can see the locations of two settlements at the foot of the
temple; the eastern settlement was oval shaped and badly eroded away, whilst the
western settlement was a well planned settlement and better preserved. The
Arnold’s had only two days to do a test dig on the western settlement, which
enabled them to discover well preserved low layers of brick walls and ceramics
which belonged to the 12th dynasty. Subsequent excavations of the settlement sites
by the Institute of Archaeology of the Jagellonian University in collaboration with
8
Adapted from fig 1, ‘The Middle Kingdom Settlement at Qasr el-Sagha, 1979 -1988, Joachim Sliwa
6
the German Archaeological Institute, would confirm the initial findings by the
Arnold’s.9
On the settlements Joachim Śliwa would say;
“They were parted one from another by a deep wadi and each had its own
fortification system, the presence of which has not even been earlier suspected. As
can be clearly ascertained on the ground of studies done in the precincts of the
better preserved Western Settlement, these were the structures erected under the
supervision of architects, with their rooms based upon a regular plan and the walls
orientated along the N-S and E-W axis. The inhabitants of both these twin urban
type structures had undoubtedly been associated with the functioning of the temple
and that of the nearby necropolis (surely a classic definition of the “workers
village” can be employed here.)”10
The principal defence of the western settlement, were 2.1m thick brick walls, set in
a foundation trench some 1.10 – 1.50m deep; it was also noted that important
points, such as the corners and gates into the settlement were reinforced with
stone.11 The wavy mud brick walls that we see in Middle Kingdom sites, also
makes a strange appearance here, in the form of a single course and only one layer,
forming a wavy line on the inside of the wall: mostly against the west wall, though
some was found by the south gate. Śliwa would suggest that the wall may have
been connected with some kind of foundation ritual.12
The impressive defences and the somewhat austere style of the temple might be a
product of its somewhat remote location, and protection against possible libyain
raids. The excavations would conclude that the oldest stages of the settlement
would date to the 12th dynasty; the youngest to the 13th dynasty, and it is thought
that the abandonment of the settlements may have been at the close of the Second
Intermediate Period.13
Why this location was chosen is not known; though prior to the Middle Kingdom,
an ‘L’ shaped quay, dating to the old kingdom, approximately 1km south west of
the temple, connected to a road that led to the Widan el-Faras quarry. This was an
9
Qsar el-Sagha, studies on the Middle Kingdom and the Second Intermediate Period Settlements in 1979-1985.
Fontes Archaeological Posnanienses Vol. 36, pages 189 – 216. J.Sliwa
10
Ibid, page 191
11
Ibid, page 192-193
12
Ibid, page 206
13
Ibid, page 192
7
11 km long paved road that would connect with the basalt quarry; this quarry was
predominately used by the 4th & 5th dynasties of the old kingdom, who would use
significantly large quantities of basalt for some temple pavements, amongst other
items. Pottery finds at the quarry date to the 4th & 5th dynasties, also found at the
quarry was Early Roman Period pottery.14 Ian Shaw would state;
“The site at Qasr el-Sagha appears to be linked by an ancient paved road with the
basalt quarries of Gebel Qatrani, about 6 miles (10 kilometers) to the north. The
obvious assumption is that the settlements and temple relate directly to the
quarries exploitation, but it has been pointed out by a number of scholars that
basalt was mainly used in the Old Kingdom and Late Period, and indeed the
pottery associated with the quarry road itself is primarily Old Kingdom in date. It
is therefore something of a mystery as to why the major features of archaeological
remains at Qsar el-Sagha should date to the twelfth dynasty, when very little basalt
was being used. Perhaps Qasr el-Sagha originated as a very small quarry-workers
settlement. There are as yet no traces of Old Kingdom activity at the site, but by
the Middle Kingdom it had evidently developed into a different kind of community
with some as yet unknown function, perhaps connected with the empty, unfinished
temple.”15
The excavations by the settlements and the temple appear to show no trace of old
kingdom presence in these areas at least. Śliwa would suggest that the western
settlement performed housing functions, whilst the eastern settlement appears to
have been an area of production; here, fragments of unfinished artifacts, such as
stone vessels and statuettes were exclusively found, along with some tools.16 Śliwa
would also suggest that the rich mineral deposits in the area may have been a factor
in site location; items such as Limestone, gypsum, basalt, dolerite and Petrified
wood.17 Ultimately the mystery of the site will remain, but whatever the rationale
in its location, significant resources were allocated to its construction. Yes, the
temple may appear small compared to impressive New Kingdom temples, but
craftsmanship is impressive, suggesting an important role in this location.
14
Old Kingdom Basalt Quarrying activities at Widan el-Faras, Northern Fayum Desert. E.Bloxam & P.Storemyr, JEA
Volume 88, 2002, page 26
15
Exploring Ancient Egypt, Ian Shaw, 2003, page 114
16
Qsar el-Sagha, studies on the Middle Kingdom and the Second Intermediate Period Settlements in 1979-1985.
Fontes Archaeological Posnanienses Vol. 36, page 192
17
The Middle Kingdom Settlement at Qsar el-Sagha, 1979-1988, Prave Archeologiczne Z.54 – 1992, J. Sliwa, page
20
8
The Temple
I am most grateful once again to the Isida Project and Olga Kozlova for the use of
their excellent images.
The front of the temple is largely dressed down, with only a few bosses still
evident; the sides and rear of the temple still maintain excesses of stock that had
yet to be dressed down: the back wall had its dressing interrupted, with only the
upper course and a portion of the middle being dressed, before work ceased.18 The
largest blocks are installed in the temple front, whilst the sides and rear largely
contain smaller blocks; this provides a mostly unbroken canvas for those tasked
with the temple decoration, which was never started.
Overleaf, the top image shows the temples west wall, whilst the lower image
shows the temples east wall.
18
Der Tempel Qasr el-Sagha, 1979, Dieter & Dorothea Arnold, page 17
9
10
The above view shows the north and east walls of the temple. Unlike the western
settlement which appears fairly well aligned to the cardinal points, the temple is
more orientated south west to north east: I am not aware of any archaeoastronomy
studies being carried out on the temple, but it might be worthwhile to see if the
temple had any possible astronomical alignments.
As can be seen from the images, only the front façade of the temple contains large
orthostats; Dieter Arnold would state;
“This large stone façade has given the temple a date in the Old Kingdom.
However, this is not a real »megalithic« construction with giant blocks extending
through the walls, but the walls were clad with technical sophistication with
orthostats, which are only supposed to simulate a »megalithic« character.”19
19
Ibid, page 17
11
As for the oblique joints found in areas of the construction; Arnold would state:
“These are found not only in the Old Kingdom but in pharaonic buildings of all
times, so they are also no argument for dating the building to the 4 th dynasty.”20
Dorothea Arnold who examined the ceramics found at the temple and in its vicinity
could find no fragments from the Old Kingdom. The fragments that were found
from the clearing of the temple and nearby came from early Ptolemaic, late Roman
and Islamic times: some New Kingdom fragments were found in debris directly
above the desert floor.21 Based on the ceramic finds it was suggested that the
demolition of the temple began in or after the Ptolemaic period.22
The strongest evidence for the dating of the temple comes from the actual
construction rubble of the temple, found in debris heaps south west of the temple;
here was found the oldest ceramics in the vicinity of the temple, which
corresponded to the reign of Senusret II-III.23 Dieter Arnold would favour Senusret
II, whose pyramid lay at the mouth of the Faiyum, as possible builder of the
temple.24
The Foundations
The base of the temple is cut into the tafl, with the foundations being made of three
layers. The lowest layer consists of a very crumbly tafl rock, some 37-41cm thick
and barely distinguishable from the natural tafl rock on which it lays. The middle
layer is described as an almost white sandstone some 30cm thick, and finally the
top layer, the thickest at some 51-57cm thick, made of a very hard limestone.
These substantial foundations were enhanced under the rear wall of the temple, by
a fourth layer due to the higher terrain level in this location.25
These three foundation layers appear to run throughout the temple, except for the
long offering hall in front of the seven shrines; here, the lowest layer is replaced
with sand fill. The side rooms accessed at both ends of the long offering hall are
thought to have had three layers, but the actions of searchers have badly damaged
the foundations in these locations to be sure. Two circular shafts were found under
the temples south east and north east corners, cut in the tafl under the broken
20
Ibid, page 20-21
Ibid, page 20
22
Ibid, page 18
23
Ibid, page 20
24
Ibid,,page 21
25
Ibid , page 9
21
12
foundations; though it was not possible to determine if these were foundation pits
or activity of searchers. A search was conducted in the south west corner, but no
foundation pit was found; a search could not be conducted in the north west corner
as a large displaced stone beam prevented access. Another possible foundation pit
was found in front of the entrance, which was partially walled with brick.
I am grateful to Dieter Arnold for his kind permission to reproduce some of his
images. The above image26 shows the possible foundation pit in front of the temple
entrance. Searchers have made considerable excavations through the foundation
layers in many locations, to such an extent that they have practically undermined
the temple; Arnold was concerned that the rear wall may have tipped over in the
near future, and so filled many of the voids in the foundation with sand.27 Searcher
activities combined with possible seismic events, weather events, along with stone
robbers, have left us a structure, badly shaken, with many large joint openings, but
still standing.
26
27
Ibid, plate 8a
Ibid, page 9
13
14
15
Page 14, top image, taken from the temple entrance, shows significant damage to
the central shrine.
Page 14, lower image, and page 15, top image, highlight masonry movement.
Page 15, lower image, highlights the precarious hold of the temples rear wall.
Reproduced with permission of the Griffith Institute, University of Oxford28
The above image by Petrie shows the rear of the temple, with stone debris and
dislodged blocks. The handling boss, middle right, can also be seen on the lower
image, page 15.
28
http://www.griffith.ox.ac.uk/gri/ppoe/Petrie_5_8_030lower.html
16
The above section of the temple, taken through the entrance axis, gives us a rough
idea of the temple foundations
The Entrance
The entrance to the temple is accurately placed in the middle of the temples south
face, judging from Petrie’s measures (see plate VI, page 5); the entrance door
frames to their respective corner’s, differ by only 0.3 of an inch (398.1 versus
398.4 inches). Petrie gives the width of the outer doorframe as some 48 inches,
giving us a total length of 844.5 inches (21.45m). The entrance door posts have a
convex profile, which would further reduce the entrance width; beyond the door
posts the entrance widens to 57 inches, for a distance of 49.8 inches. The door
17
posts at 23 deep, suggest a thickness of the front wall of some 72.8 inches, or a
possible 3.5 cubits.
In this view of the entrance one can just make out the curvature of the door posts.
After the door posts, the passage widens to accommodate the temple door; the
single door pivoted on the left and would have been secured by a bolt, the hole of
which is visible above, immediately behind the right door post. The convex door
posts whilst a feature in other Middle Kingdom structures (for example, we have
three pairs in the burial chamber of Senwosret III29) is also to be found in Djoser’s
step pyramid complex, see image overleaf.
29
The Pyramid Complex of Senwosret III at dahshur, Architectural Studies. 2002, D.Arnold, plate 15
18
Image courtesy of Greg Slater
The above image is from one of the chapels that flank Djoser’s Heb-Sed court, and
one of the earliest examples of the convex door post.
19
In the above view, from inside the temple, we can see that the entrance is spanned
by three large stones, which have been dressed underneath to provide a horizontal
door stop. The pivot hole for the door is visible in the corner. Any door fitted in
this location will have to accommodate most of the 57 inch width of the entrance;
but as the rebate for the door is only 49.8 deep, this means that any door in the
fully open position will intrude into the offering hall.
Entry to the temple appears to be a two stage process; in order to open the main
door, one must go through another smaller doorway in the front façade; this
smaller doorway leads to a long narrow corridor, some 20 inches wide, which
terminates near the entrance (see Petrie’s plate VI, page 5). At the end of the
corridor, a cutting has been made in the side wall, just below the ceiling; this
would provide access to the bolt, which locked the main temple door.
20
Image courtesy of Colin Reader
At the end of the narrow corridor a niche has been made in the side wall just below
the ceiling. In the image above we can see light coming through the bolt hole (the
other side of the hole can be seen on the image on page 18). According to Arnold’s
scale drawings30, the round hole is about 19.0cm long, diameter 9.0cm.
Immediately behind the round hole we have a neatly cut rectangular portion 43cm
long, 23 cm wide and 17 cm high. Arnold suggested that when the temple was
locked, a stone would be inserted into the niche to hide the holes location.
The floor of the long corridor appears to coincide with the middle foundation layer
of white sandstone, this gives a corridor height of about 2.1m31 (4 cubits?); this
would make the lower ledge of the niche about 1.72m above the floor, high enough
to avert prying eyes, especially in a dark narrow corridor. Maybe those tasked with
opening the door brought a stool along with a lamp. This somewhat elaborate
security made Arnold suggest; “the temple was not used and guarded all the time,
but was left to its own devices at times and exposed to the risk of looting. The
30
31
Der Tempel Qasr el-Sagha, 1979, Dieter & Dorothea Arnold, page 13, fig 7.
Ibid, page 12, fig 6.
21
nearby settlement was therefore not permanently inhabited.”32 While this may
have been the case, it would seem strange to build such an impressive temple for a
settlement that may have seen only seasonal use. The impressive defensive walls of
the settlement, together with the somewhat austere fortress like appearance of the
temple, might be indicative of its remote location. The location may not have been
a secure one; as darkness fell, workers and priests alike would retire to the well
defended settlement, after securely locking the temple against any possible
marauders.
Arnold’s reconstruction shows a simple bolt being pushed through the hole and
engaging into hoops fastened to the back of the door to secure it shut. However, the
neatly made rectangular area behind the hole makes me wonder if something more
complex could have been used. It might be possible that this rectangular space held
a wooden box, which operated the bolt via a variation of the Egyptian pin lock.
Such locks are still a mystery, many sources, including Wikipedia, suggest that
such locks date as far back as 2000BC in Egypt; though I have seen articles that
question this. Petrie would mention a little about locks in his „Tools and Weapons‟
publication, 1916; and I have shown his plate LXXV on the next page that shows
some examples of locks and the keys. These wooden keys would come in all
shapes and sizes, often over a foot long; certainly the niche could accommodate
such a key. Ultimately we may never know if such a lock fitted into this hole, but I
thought it best to alert the reader to the possibility.
The doorway to the narrow corridor may have been intended to also have had its
own door, as Arnold found a pivot hole in the ceiling, though parts of the wall had
not been dressed enough to allow a door leaf. Arnold suggested that the idea of a
door may have been abandoned to be replaced with two large stone blocks;33
though it might be that the incomplete nature of this door is just down to the
unfinished nature of the temple; work ceased and the temple abandoned during its
creation. It cannot be discounted that this smaller door may have had its own lock
for security.
32
33
Ibid, page 13
Ibid, page 13
22
23
The above section gives us a rough idea of the corridor inside the front façade.
Petrie gives the narrow corridor’s entrance as 74 inches from the south east corner;
a possible 3.5 cubits, to match the thickness of the front wall. If we take the
entrance and corridor to be an intended 1 cubit wide, then entrance axis would be 4
cubits from the corner. The small entrance axis appears to align with the axis of the
eastern room; this room Petrie gives as 213.6 long by 83.6 inches wide; if we take
the architects intention that the room is 4 cubits wide, with the temples exterior
east wall as 2 cubits thick, then the axis of the room and entrance will align.
Images overleaf; top - view of corridor entrance. Bottom left (Arnold’s plate 12a) view along narrow corridor. Bottom right (Arnold’s plate 12c) - view along
corridor towards entrance: here we have cuttings into the east wall; it’s hard to say
if these are searcher activities, or purposely cut to provide storage spaces. The floor
at the end of the corridor near the main entrance had been excavated into by
searchers and reaching into the long offering hall.
24
25
The Eastern Room
Above we have the entrance into the eastern room; the block that I have
highlighted, supporting the door lintel, is one piece that would continue into the
south façade and roof the narrow corridor. Just visible on the lintel is a slight
projection that would frame the doorway; such projections can also be seen on the
interior side of the main entrance, and the doorway that leads to the western rooms.
26
In this view of the eastern doorway taken further back along the offering hall, I
have highlighted the blocks that would roof the narrow corridor; this narrow band
is also visible along the front façade. It is interesting to note that the large
orthostats that we see on the front façade are also replicated on the interior side.
By scale rule from Arnolds’s plate 25 & 23, the height of the doorway is about
2.63m (5 cubits?), and the width 1.05m (2 cubits?). The span of the large door
lintel also appears to match the height of the door at 5 cubits; so the width of the
long offering hall is 5 cubits. Left door jamb 2 cubits wide, doorway 2 cubits wide
and right door jamb 1 cubit wide (Petrie appears to have made an error on his plate
VI as he has the doorway at 31 wide, with the left jamb as 42 inches wide; but the
images rather suggest that they are the same width).
The top of the lintel is about 3.85m above the floor.
27
This view from inside the eastern room looking towards its entrance; the red
highlighted area is the approximate position of the narrow corridor entrance,
behind the orthostat. Petrie would give the length of this room as some 213.6
inches, though it is not known if he took into account the significant opening of the
wall joints, present in the room: it is possible that the room was to be an intended
10 cubits long.
The pivot hole for the door is on the south end of the lintel, so when fully open, it
would not obstruct access to the room, but stored close to the south wall. The
doorway from the offering hall side is 5 cubits high by 2 cubits wide, though the
actual door will be slightly larger to accommodate the door stops, visible from
inside the room. In the images I have, I see no evidence of holes in the opposing
door jambs that could accommodate a bolt to secure the door.
28
This view looking at the rear wall of the eastern room, we see what appears to be
another doorway, the threshold for this opening is about 1m above the floor.
Compare this image with the image on page 11, and one will see that this stone is
part of the upper foundation layer for the rear of the temple wall. To the right of
the opening on the inside a rough cutting of some 1.2m wide has been made, as if
to accommodate a door. Arnold was not convinced that such a door was planned
from the beginning, but rather a later breakthrough; as it would seem inconsistent
with the elaborate security provided for the main temple entrance.34
34
Ibid, page 12
29
The Western Rooms
At the other end of the offering hall, we have the doorway into the western rooms;
both the western and eastern doorways are accurately placed from the main temple
entrance (Petrie gives 237.5 from main entrance to west doorway, and 238.0 inches
to east doorway). Like the eastern doorway, we can see a flat projection framing
the doorway. The western doorway has a much larger lintel spanning the door, on
the reverse near the top a shelf had been cut to accommodate ceiling beams, which
gave the western rooms a similar height to the eastern room.
Interestingly the western doorway is smaller than the eastern; its height from
Arnolds plate 23, is only 2.1m (4 cubits, compared to the eastern 5 cubits). Petrie,
gives the doorway as 35.2 wide, the left jamb as 21, and the right jamb as 50.4, for
a total of 106.6 inches. Arnold would develop a cubit grid plan for the temple35 that
shows the width of the western door to match the eastern at 2 cubits (41.3 inches),
however, whereas Petrie’s measure appears wrong for the eastern door, the images
35
Ibid, plate 27
30
suggest that he is correct for the western door, and that therefore this smaller
doorway does not fit Arnold’s cubit plan.
This view from inside the first western room shows the neatly made doorway; here
the pivot hole for the single leaf door is opposite to that of the eastern doorway, in
being to the north side of the lintel. The rebate for the door Petrie gives as 13.4
inches long; the door stops combined add another 8 inches to the width of the door,
so we have a possible door of about 43 inches wide (35.2 + 8); so an open door
would protrude into the room by 29.6 inches (0.75m).
If one looks at Petrie’s plate VI (page 5) it will be seen that the thickness of the
front façade has been reduced to provide more space for this room. Petrie gives
116.3 by 85 inches for the room (5.5 by 4 cubits?). The wall with the strange shelf
cutout he gives as 38.4 wide, with the wall the other side of the entrance slightly
less at 34.5 inches; subtracting these from 116.3, suggests a rebated door width of
43.4 inches. From the images I hold, there appears to be no hole in the door jambs
to secure any door.
31
The western rooms consist of two compartments; to access the rear room, one has
to crawl through an opening in the dividing wall, some 75 cm high by 53 cm (1
cubit) wide. The rear compartment above, with the figure inside, according to
Petrie is 111.0 inches E-W, and 90 inches N-S. Its east wall is shared with the
western most shrine, and a bench runs along the base of the wall, this bench
matches the entrance height, and appears to be a continuation of the platform on
which the shrines sit, as this is also 75cm: the bench may have been used for
storage. Arnold would suggest that the temples more valuable items would be
stored in this compartment. How the small entrance was concealed and the role of
the strange shelf cut in the wall will remain unknown; but it is interesting how they
hung the door on this side, for in the open position the door would obscure the
small opening: possibly behind the open door, a wooden chest could act as a
continuation of the door stop and cover the opening, with the shelf containing a
few items.
32
The entrance to the rear compartment, the highlighted block forms part of the
bench in the rear compartment
33
The Offering Hall
The long offering hall contains the seven shrines; the above image shows six of
them. The shrines sit on a raised platform some 75cm high; a small projection runs
along the length of the platform some 14cm high and deep. This creates a shelf in
front of the shrine walls 1 cubit deep, therefore the distance between the shelf and
the rear of the façade is 4 cubits, whilst the walls of the shrine to the rear of the
façade is 5 cubits. The middle shrine opposite the main entrance is wider than the
other six, Petrie gives 67.8 inches (1.72m); the remaining shrines appear to be 2
cubits wide, with the mean of Petrie’s measures being 41.7 inches. However, when
one looks at Petrie’s measures (see plate VI, page 5) a strange pattern emerges, in
that the shrines ascend in width towards the central shrine, which may be just a
coincidence. Arnold would state;
“What is striking in the area of the chapels is both the accuracy of the dimensions,
the care taken with the stone processing and the selection of particularly hard and
regular stones.”36
36
Ibid, page 11-12
34
The above image of Medinet Madi, shows the similarity of design that we see at
Qsar el-Sagha. Here, the larger central shrine contained a stone altar that may have
contained three statues; note the surviving texts that decorate the walls. We may
never know if a similar setup was envisioned at Qsar; the central shrine there,
being especially damaged by searchers. Madi held only three shrines whilst Qsar
had seven.
Due to the greater width of the central shrine, it was necessary to place the double
leaf doors, which opened outwards, further back into the shrine; so that in the open
position the widened rebates in front of the shrine accommodated the doors,
without them encroaching into the shelf area.
35
The above view is looking into the westernmost shrine; the pivot holes for the
doors are clearly visible. The finish of the walls appears quite fine, awaiting their
decoration, which never started.
36
Above, we see the architect’s solution to the decorative features that frame the
shrines. Below, all temple doors have a cutting in the floor, to allow door fitment.
37
Arnold would give the height of the shrine doors as 5 cubits; the tallest door is the
main temple entrance, which by scale rule appears to be 6 cubits high.
On Arnold’s plate 27c, he created a design scheme for the shrines which I have
replicated above; the above being a grouping of three of the smaller shrines.
38
The above plan scheme is reproduced from Arnold’s scheme on plate 27b
Arnold would develop the above scheme for the length of the offering hall; red
numbers = cubits, white numbers = palms; total length being 25 cubits 5 palms.
39
The central shrine at 3 cubits wide requires that its doors are set back, in relation to
the others, by a further half cubit. Arnold’s scheme of 25 cubits 5 palms would
seem strange, given the care that the architect has given to whole cubit distances
elsewhere in the offering hall. For example, rear wall of shrines to rear wall of
front façade, 10 cubits: rear wall of shrines to front wall of shrines, 5 cubits, and to
projection of platform 6 cubits, From the front wall of the shrines to the rear of the
front façade 5 cubits; so we might logically ask why the E-W length of the offering
hall is not a round 25 cubits.
Arnold provides no detailed
dimensions in his work, so we
have to rely on his scale
drawings
and
cubit
reconstructions.
The
only
dimensions we have are from
Petrie’s plate VI; here the
tolerances appear large, with the
2 cubit wide shrines, ranging
from a low of 39.8 inches to a
high of 43.3. Likewise the
central shrine he gives as 67.8
wide, which seems in excess of 3
cubits; while across the offering
hall the entrance at 57 wide
appears too low for 3 cubits.
Clearly the structure has been
badly shaken and can have an
effect
on
measures.
I
experimented with many of the
images I have to see if there
could be a solution: in the image above, we can see that the decorative bar, which
Arnold gives as 1 palm wide, appears to fill the space either side when replicated
three times. I tried this on several images and obtained the same result, so it may
be possible that this width was intended to be 1 cubit wide instead of Arnold’s 1
cubit, 2 digits. By adjusting this one dimension onto Arnold’s scheme, we obtain a
possible new scheme overleaf, which provides us with 25 cubits. Of course such a
scheme can only be tentative, as detailed measures allowing for masonry
movement would be needed to clarify what the architect intended.
40
A possible scheme for the offering hall, which provides a 25 cubit solution
Arnold suggested that the temple’s foot print was an intended 15.5 by 40 cubits 37. I
have done a tentative reconstruction above, based on the undressed exterior walls
being an intended 2 cubits thick. Architects plans are always subject to the
intentions and quality of the masons tasked with the job in hand; they have to work
with the stone supplied and any possible limitations incurred with the material at
hand, necessitating slight amendments to the architects plans: in the case of the
badly damaged and unfinished temple, one can only offer a best guess on the
architects intentions.
37
Ibid, page 17
41
The Doors
As one can well imagine, doors are a rare
commodity from ancient Egypt; destined
to be repurposed for other uses. The door
left is a rare survivor from the Old
Kingdom38. All shrine doors were double
leaf, with the remaining doors being
single leaf; how these doors may have
been decorated is unknown. Petrie in his
Egyptian Architecture publication would
state;
“The method of pivoting was by the post
projecting into a cup-shaped pivot-hole
below, and through a cylindrical hole in
the lintel. The bearing on the cup was
usually too shallow for the thrust of the
door weight, and made the pivot beam
kick out against the jamb. Doors that
could be handled were usually set up by
skewing, so that the top of the post could
go up into the lintel, and. then by
dropping the door pivot into the cup
below: this is well seen in the threshold
of Pepy (Abydos II, liii). In the granite
temple of Khafra, the pivot block is flat
polished black basalt; how the thrust of
the door was maintained is not visible.
Conical pivot holes were used from the
iind dynasty or earlier (Hierakon.
iii,36).39
Image Courtesy of Greg Slater
38
39
th
JE 47749, Cairo museum, 6 dynasty, Saqqara
Egyptian Architecture, Petrie, 1938, pages 74-75
42
To assist in fitting the doors, a channel was cut in the floor (see bottom image on
page 37); this channel allows the door to be introduced, then raised so that the
upper door pivot engages into the lintel hole; the bottom pan would be introduced
and the door lowered slightly onto the pan: the remaining channel would then be
filled, with masonry. Amongst all the temple doors it appears no pans were found,
and all the channels appear empty; it might be the case that the builders never got
to the stage of fitting the doors, as the temple decoration had yet to commence, and
doors at this stage may have been a hindrance.
For the main temple entrance, Arnold would give the lintel pivot hole as 12cm
deep and 10 cm in diameter; its accompanying floor channel is given as 10cm deep
and 14cm wide. Arnold would state that the channels ran perpendicular to the
direction of the passage; a common feature in the 12 th dynasty, and he was not
aware of any older examples.40 In other dynasties we sometimes see this channel
run parallel to the passage walls, sometimes sloped, to assist door fitment.
The Roof
The only surviving part of the roof covers the seven shrines; these roofing stones
being still protected by a row of masonry, laid on top of their ends. The stone
robbers have successfully removed the remainder of the roof, along with upper
wall courses on the temples narrow ends. Architectural traces found by Arnold on
the upper stones, indicated that the roof beams had been fitted. Arnold would also
add;
“However, the fact that the ceiling already existed must be concluded from the fact
that the interior of the temple and the façade were already completely smooth and
prepared to receive the decoration, i.e. that a construction phase had been reached
during which the temple roof had already been laid on.”41
Arnold would suggest that these roofing beams would be brought up via ramps,
placed at the narrow ends of the temple. Some of the upper stones showed channels
for the insertion of a wooden runner to assist in moving the stones. There is
nothing in the report to indicate the remains of any ramps, which might be seen as
further evidence that the roof had been completed and the ramps removed; for if
the temple roof was halted halfway through construction we might expect to see
40
Der Tempel Qasr el-Sagha, 1979, Dieter & Dorothea Arnold, page 11.
Ibid, page 14
41
43
evidence of these ramps still in situ, as it’s hard to believe that people took care to
dismantle the ramps after the decision was made to abandon the site.
Image Courtesy of Francois Oliver, www.meretsegerbooks.com
44
The image on the previous page shows the surviving roof, which covers the seven
shrines, looking west. Though this corridor like space would not be seen from the
ground, the builders still took care to dress the faces: Arnold found it striking that
no drainage channel was provided, the layout of the roofs suggesting that one
should have been in the middle of the rear wall.42 Found amongst the dislodged
masonry that surrounds the temple site, were coping stones, which are thought to
have run around the perimeter of the temple roof.
Based on Arnold’s reconstruction, the yellow elements of the roof have been
removed by the stone robbers. Temple illumination is not known; Arnold
suggested a possible light slit on the end wall: the ceiling for the east and west
rooms was at a lower level, below the light slit, but higher than the shrine ceiling.
42
Ibid, page 15
45
Roof reconstruction
The east and west rooms may also have some variant of light slit to illuminate
these spaces.
46
If ramps were used at either narrow end of the temple to bring up the roof blocks
for the offering hall, then it’s possible that the exterior walls at the ends were
constructed last, and this may explain why we still see traces of handling bosses in
these locations. Those tasked with robbing the stone roof would probably
dismantle the end walls first, to give access to the roofing beams; again there is no
evidence of any demolition ramps to undertake the job. Arnold would note that
there was little stone debris in the offering hall, and that the roofing beams must
have been removed carefully;43 Arnold would suggest that such stone may have
been robbed to help construct the Ptolemaic temple at Dimai.44 To remove these
roofing stones the robbers may have used large wooden baulks, as a makeshift
ramp to lower them. Fortunately the robbing operation appears to be short lived, as
the vast bulk of the temple still remains.
Also in the image above, we can see that the upper foundation layer that supports
the wall stones of the long corridor, are noticeably thicker; the masons have
dressed down this excess, such that an upper portion of the foundation stone is
incorporated into the façade wall.
43
44
Ibid, page 14
Ibid, page 19
47
The Stones
The stones that make up the vast bulk of the temple are described as Upper Eocene
limestone, which contain a lot of fossils and sand, whose hardness is about 4.5 on
the Mohs scale.45 An onsite inspection by two geologists could not determine the
origin of this limestone; though it is thought that the sandstone and tafl foundation
layers could be from the local area. It would be most interesting to trace the origin
of the temples fine limestone.
This view looking into the north west corner of the offering hall; we can see a part
of the flat door projection, which frames the western rooms doorway: the sand
filled door channel for the shrine doors is also visible.
45
Ibid, page 16
48
This view is of the first western room, doorway on left; one can see how some
blocks appear to turn a corner, as the masons removed excess stock; this is quite
prevalent throughout the temple, and typical Egyptian masonry technique. The
joints, where not disrupted by temple movement, are very fine and highlight the
fine quality of the temple.
49
This view from inside the eastern room highlights some of the masonry movement;
here we can see through to the offering hall and the decorative feature that frames
the shrine on the other side of the wall.
50
Arranged around the temple we have groupings of masonry from the site. In front
of the temple we have one of the larger groupings, forming a sort of crescent
shape; here large amounts of stone slabs appear to be stacked together. Some of
this masonry was found stored on top of debris, which had to have been present
before the storing of the slabs, and ceramics found amongst this debris were dated
to the Ptolemaic period. Arnold would state;
“It must therefore be concluded with certainty that the demolition of the temple
began in or after the Ptolemaic period – some isolated blocks lie about 120m down
the valley below where they were left during transport. For unknown reasons, this
work was stopped in time before the temple was really destroyed.”46
It is thought that a lot of these slabs are the remains of a pavement that surrounded
the temple, which were subsequently uprooted and awaiting transport to possibly
Dimai. Arnold also found similar coping stones to those found on the temple, but
these displayed an inclination on one side, and vertical on the other, whereas the
temple walls are vertical: it is thought therefore that they may have formed part of
a wall, though no foundation trenches were found for any wall.
46
Ibid, page 18
51
View of stacked slabs, and in the back ground coping stones. Below, dovetail in
one of the dislodged blocks; such blocks are found on the temple corners.
52
Concluding Remarks
Without doubt the temple has taken a bit of a battering, a mixture of natural events
and human destruction; but enough remains to show that a lot of care and resources
were allocated to its construction in this remote location. The reason for its
construction and the role of the nearby settlements may always largely remain a
mystery; but the unfinished nature of the temple might suggest a change of ruler,
who no longer cared for this project, and quite happy to abandon the site.
What was proposed to be placed inside the shrines will also be a mystery, though
Sobek is one of the likely candidates. Can we be even sure that the temple was in
use? It’s possible that it was always just a building site, the shrines empty, left
open with no doors. Would they place their important gods inside these shrines,
while so much work remained to be done?
One can’t help but think that yet more is to be discovered from this temple; Arnold
had only a month and some of this disrupted by serious storms, but the good work
that the Arnold’s have done, should leave no one in any doubt that this is a fine
Middle Kingdom temple, and maybe one that should be better preserved and
protected: granted its remote location is difficult to police. Its remote location has
largely saved it from the worst of human depredation; but at the same time, out of
sight, out of mind, means that it can be too easy to forget its existence and hamper
further research. Let’s hope it receives more attention in the not too distant future.
53