The earliest monumental structures in Egypt have been dubbed `fortresses' - although their true finction has not been established. Surveying these enigmatic monuments, Eric Uphill proposes that they are in fact the royal Ka Houses of Egypt, the development of which reaches its peak in the Middle Kingdom. ERIC P. Origins The two `Shuneh' enclosures at Abydos dated to the 2nd Dynasty have never been satisfactorily explained. The old designation of fort is valid only in so far as their outer wall defences appear to be a military feature, while the more recent designation of `funerary palaces' given by Kemp, although preferable, is only applicable if seen in the context of structures erected in a cemetery area but with no actual burials associated with them to date. As an extension of funerary structures built elsewhere their function would make better sense, and it is the purpose of this article to attempt to define this role more clearly. The earlier complex dated to Peribsen is much less substantial than the later one, measuring some 99 m N-S x 54.9 m E-W without its missing north wall, and thus covering just over 1.4 acres. With its north wall, it would thus have been at least lOOm long, as its single enclosure wall is only 1.2m thick at the base according to the excavator (the plan showing it varies between 1.25 m and 1.5 m thick). This includes the orna- mental exterior pilasters which are a feature of these early structures. The wall was thus only about 2.5 cubits broad and originally stood 4.6m high, being therefore hardly of a real defen- sive nature. Allowing for this wall thickness, the interior width of the enclosure is of the order of 52 to 52.4m - that is 100 cubits exactly. But for one small interior mud brick building, possibly a wine store judging from the jar JACF VOL. 5 (1991/92) UPHILL sealings it contained, there is nothing to indicate the purpose of the complex, whether funerary or otherwise. The later complex, belonging to Khasekhemui, is a much more formidable affair. With double walls it measures about 121 m N-S x 53.9m E-W within its inner wall.2 This inner circuit wall is 5 m thick at the base and still stood 11 m high in places in Ayrton's day, so he estimated its original height at about 12m. The enclosure was then around 123 m x 64.6 m overall. The outer wall was also substantially greater than that of Peribsen, averaging 2.4m thick and standing from 5.5 to 6m high, being probably not much taller originally. Including the blank space between the two walls, the layout extended to roughly 134m 0 50 I I METRES /" Above: The Abydos enclosure of Peribsen; Right: The Aby- dos enclosure of Khasekhemui (known as Shunet el-Zebib). [Illust: H. Jaeschke, after Ayrton et al., op. cit. [I], pls. VII, v & VI] Pages 77-88 77 View along the east passage between the outer and inner walls of the Shunet el-Zebib at Abydos. [Photo: D. Rohl] N-S x 77m E-W or about 2.5 acres. The inner rampart was also decorated with pilasters on the exterior face, those along the west wall being 53 cms average breadth, that is 1 cubit. Gateways in this wall were of the fortified type featuring indirect and `dog's leg' approaches for safety. Once again the only internal structure was a brick storehouse building in which sealings of the king were found. Even allowing for a slight batter or inclination on the walls, the volume of materials used is very considerable, up to 22,280m3 for the inner rampart and 5960m3 for the outer, the total of 28,240 cubic metres being equal to an average 12th Dynasty Nubian fort or nearly one million cubic feet of brickwork. It marks a vast increase on the approximately 2,380m3 of the Peribsen enclosure. This is well over ten times the amount of the earlier structure and at least eight that of the largest 1st Dynasty funerary enclosure at Abydos belonging to Djer's reign. Another similar enclosure of Khasekhemui exists at Hierakonpolis and, although covering less area, it is equally strongly fortified. It measures about 57m N-S x 47.5m E-W internally,3 probably 110 x 90 cubits as planned. Its inner surrounding wall is 4.87m thick at the base, giving a total length of around 66.74m x 57.24m breadth. Beyond this is a narrow corridor or blank area 2.25 m across, then an outer wall measuring 2.34m thick at the base, the whole layout thus reaching 76m N-S x 66m E-W (250 x 217ft.) or 1.25 acres. A major difference to the Abydos enclosures lies in it having a single great south- eastern gate which is flanked by two massive 78 towers of 5.25m or 10 cubits pro- jection and 6m breadth across the face. The wall has pilasters on the exterior and was plastered like the Abydos ramparts. A granite door- post inscribed with the name of Khasekhemui, found just outside the gateway, is both large and decorated with reliefs, as illustrated by the Boston expedition. This may well indicate that the enclosure contained some kind of shrine or chapel (if it did not come from the nearby town, which seems unlikely given its prov- enan~e).~ Remains of a brick build- ing were found a little to the south west of the centre of the fort but it was much too destroyed to be fully planned. In addition two granite blocks were recovered by the later English expedition (one being a column base) which also suggest that Lansing's doorpost may have belonged to the same building. In this context it should also be noted that the two famous Khasekhem statues now in Oxford and Cairo were also found here in the nearby temple area (one of limestone5 and the other of schist6) with details of the king's wars carved on their bases, perhaps suggesting that a personal cult of the king once existed here. Allowing a height of up to 10.5 m or 20 cubits for the inner wall and 6m for the outer (as with the Abydos enclosure), a figure of approximately 0 40 The' Hierakonpolis enclosure of Khasekhemui of the 2nd Dynasty. [Illust: D. Rohl, after Quibell & Green: Hierakonpolis I1 (1902), pl. LXXIV] JACF VOL. 5 11,395m3 + 3,905m3 or a total of 15,300m3 of material was used in its construction. This, while not much more than half that contained in the Abydos enclosure of Khasekhemui, is still over six times that of Peribsen, giving some indication of the scale on which the later builders worked. The suggestion current among some Egyptolo- gists that these enclosures represent some kind of valley or approach building - adjuncts in fact to the royal tombs sited further into the desert at Abydos - rests on their association with the tombs of the 1st and 2nd Dynasty period found at the same site. In fact, while the reigns of Peribsen and Khasekhemui are both represented among the Umm el-Kaab (Early Dynastic Abydos) tombs, there is as yet nothing to prove for certain that these are the actual burial places of the kings themselves. Indeed they could equally well belong to queens or other members of the royal family. The same argument also applies to the 1st Dynasty enclosures being re-excavated at Abydos; bar the subsidiary graves surrounding them, they may not contain a central burial place at all. So that while the sacrificial interments are a factor not present in the 2nd Dynasty enclosures, they are still not conclusive as evidence of a king's tomb complex. In the case of Khasekhemui the question becomes completely hypothetical when the Hierakonpolis fortified enclosure is taken into account. This can scarcely be a valley building or some kind of funerary temple attached to a possible king's burial at Abydos, and the search for an explana- tion must consequently be widened. Again, the strong possibility of a shrine and even statues for the king's cult being also established at the southern city, must surely indicate that much more general royal rites were celebrated throughout the country than has been previously postulated - and not just for Abydos and Sakkara. In fact, some- thing along the lines of a series of royal `houses' in all the nomes of Egypt, for use during the king's lifetime, with possibly chapels and shrines for the royal cult to be celebrated then and after death, seems a better and more likely explanation of this phenomenon. Old Kingdom Development It is during the great days of the Old Kingdom that the full range of these royal foundations can best be seen. In the Valley Temple of the Bent Pyramid of king Sneferu at Dahshur, Fakhry found evidence of a huge royal cult analogous but separate from that conducted in his pyramid complexes.7 Here rows of female figures are depicted who represent estate people gathered from nomes all over Egypt, but with some exceptions in areas where the king possibly had JACF VOL. 5 Section of the Sneferu estates list representing the nomes of the Hare and the Oryx. [After Fakhry: The Monuments of Sneferu at Dashur 11, Fig. 161 no estates. Each figure has both a hwt k3 (Ka-House) and niwt (domain) sign combined on her head, as they represent provinces with a foundation of the king, whose name is therefore written inside the house enclosure. They all carry water and bread signs as is usual in such scenes, and, while not complete, the list is very infor- mative. It should be noted that Fakhry allowed for the later total of 22 nomes for Upper Egypt, instead of the then existing 20, in his identification of provinces and places. Thus in nome IX (Akhmin) there are 3 Ka House figures, X (Qau) has 4, XI (Hypselis) 4, XI1 either 2 or 3, XIII (Assiut) 4, XIV (Cusae) 3, XV (Hare) 3, XVI (Oryx) 5, then numbers XIX-XXI are omitted, followed by XVIII (Jackal) 3, XXII also 3. There are, in addition fragments from unplaced estates. This gives a total of 34 or 35 Ka House figures in the 10 nomes thus preserved. Lower Egypt is far more fragmented: of its 16 nomes number XIII (Heliopolis) has 4 figures whilst there are about 7 others from unplaced fragments. The total is thus only 11. At a conservative estimate these figures indicate that there could be up to 100 figures representing Ka House foundations for the whole of Egypt. At this time they do not seem to be separated from the royal towns founded simultaneously in the Nile valley and Delta as later under Pepi II. More concrete evidence exists during the 6th Dynasty as to the actual form of these Ka Houses. One built by king Teti has been discovered at Bubastis. * An enclosure measuring 112m N-S x 56m E-W recalls the scale of the Shunehs at Abydos. The wall is 4m thick at the base suggesting the enclosure within was intended to be 104/5m or 200 cubits long, by 48m or 90 cubits 79 0 50 METRES The Ka House of Pepi I at Bubastis. [Illust: H. Jaeschke, after L. Habachi, op. cit. [IO], pl. 21 broad. Brick store rooms were found within, as well as a temple with limestone pillars9 on one of which there is incised a reference to the hwt k3 Tti `House of the Ka of Teti', thus giving unequivocal evidence as to its purpose. The pillars stand on square bases and belong to a brick structure. The link with the name normally given to tombs, both royal and private, in this period in Egypt is thus of more than ordinary interest and significance, as it implies an extension of the rituals normally carried out in the funerary temple attached to the king's pyramid. The parallel with Khasekhemui's buildings is almost complete, except for an outer wall circuit, but if this was much thinner it could have vanished. Another similar enclosure was added at this site by Teti's successor Pepi I. Of comparable size it measures 87.50m N-S x 64m E-W accordin while the walls are 4.70m thick ranging to 5m from the figures given by the two archaeologists. This might then suggest an enclosed area of about 79m or 150 cubits by 52.3m or 100 cubits surrounded by a wall of either 9 or 10 cubits thickness. The brick sizes were 36xl8x9cm. The to Habachi,'O or 90m x 63m after el-Sawi, 8 only known gate is sited in the centre of the south side, that is to say one of the shorter sides. The scene on the limestone lin- tel (right) recalls that of Pepi 11 at Abydos.12 The temple build- ing or sanctuary is quite sub- stantial (16.5m E-W x 15.4m N-S) with the entrance facing east. From the plan, its brick walls average about lm thick suggesting it was of no great height. It is in tripartite form with a portico containing two limestone pillars, and a hall, with eight more stone columns, 80 communicating with a row of five chapels at the rear. It should be noted that this exactly duplicates the number and, in part, the arrangement of a standard pyramid temple sanctuary with its five statue niches. The pillars reach 3.77m high or 4m with the bases, the latter being 1.12m square at the bottom tapering to 90cm across the top and measuring 47 cm high. The pillars also taper, being 70 cm square at the base and 55m at the top. All this implies a building of about 5m or 10 cubits in height above whatever base it may have had. The inner north face of the outer gate has an inscription referring to the hwt k3 Ppi `House of the Ka of Pepi' in Bubastis. Other possible Ka Houses of Pepi I may have existed at Abydos and Hierakonpolis, but, although there are certainly remains of his temple buildings at these cities, there is no definite proof as to their exact designation. The five chapel sanctuary design and statue of this king found at the latter site makes it fairly probable that here was another example of a Ka House. From these descriptions it can be seen that a fairly standard design had been evolved for such royal cult places contained within residence type enclosures. That this latter feature was also in vogue at this period in Mesopotamia can be seen by a comparison with a Sumerian palace. The structure termed Palace A at Kish dated to around 2600 BC and thus contemporary with the Old Kingdom, was possibly built by a local ruler Lugal-mu. It measures about 76 m E-W x 40.90 m N-S or a little over three quarters of an acre in area. Its outer wall is 3.50m thick and surrounds a passage 2.30m wide rather like the Abydos Shuneh's corridor or blank space; then comes an inner wall from 2.50 to 2.70m thick. Except for the inversion of thicker outer and thinner inner rampart, the similarities with Egyptian royal enclosures are striking. Pepi I1 must have built very many Ka Houses -- Lintel scene from the hwt k3 Ppi at Bubastis. [After Habachi, op. cit. JACF VOL. 5 t1019 fig. 21 f Palace A at Kish. [Illust: D. Rohl, after Mackay: A Sumerian Palace and the A Cemetery at Kish, Part II (1929), pl. XXI] during his exceptionally long reign. The lists featured on the walls of his pyramid causeway are a uniquely valuable source of information about the wide distribution of these foundations. At the upper end of the approach to the temple13 a procession of women appear as symbols of what Jequier termed the `royal castles', with geo- graphical locations, such as Cusae, written beside them in the manner of those cited for Sneferu. These are again followed by other women who represent the `royal domains' determined with a niwt or city sign, each name being commenced by names seem to duplicate the same place, the number may well be much greater than those in the first. Of this first class of buildings it can be seen that the order of size is likely to have been similar to those of Teti and Pepi I cited above, if not larger, while the second type would be of the size to be shown later. A recently discovered example of one of Pepi's Ka Houses is located as far away as Dakhla Oasis in the modem town at Ayn Asil. l6 The building was designated `d' by the excavators and really consists of two structures side by side measuring 13.50m x 8m. The principal entrance is to the north, with two wooden columns (on limestone bases 50cm in diameter) with an estimated height of 3.50m. Three rooms were placed at the south end of a court, the central one being the chapel. The whole surrounding complex was cleared over an area of 40m x 30 m showing that building `d' was bounded to the east by a wall measuring 1.20 m thick with further buildings beyond. The most important discovery was, how- ever, the actual decree of Pepi I1 in favour of the governors of the oasis of Dakhla. l7 It was found in the mudbrick building called `dI'.18 The text of the inscription, carved alongside a large serekh, reads as follows: Royal command (to) the admiral, governor of the oasis, overseer of the priests. My Majesty commands that you build a Ka House (hwt k3) in the oasis and that you engage ka-priests . . . l9 The reference thus probably relates to the funerary priests of the king - certainly to those established for a royal cult here. The temple was sited in a walled town of substantial proportions for this a cartouche esher naming Pepi OF Nefer- kare (i.e. `Spirits of Heliopolis of Pepi' or `House of the Ka of Pepi'). Jequier sug- gested that here was a representation of a series of the nomes of Upper and Lower Egypt with their farms, and that, from the preserved fragments, over 160 can be counted. In all, these drawings show four rows of Ka House figures on each side of the causeway, making at least 44 originally, l4 and a further four rows of `farm' women with well over 40 each side.15 The totals for Upper and Lower Egypt are thus: - 3 in each row, giving a possible 180, in all, e- I - 2 The Dakhla stela of Pepi II ordering the construction of a `Ka House' (hieroglyphs within box) for the king. [Illust: While a number of the second category D. Rohl, after Pantalacci, op. cit. [17], fig. 21 Ka House temple enclosures = c. 88 Town and settlement estates = c.360 JACF VOL. 5 81 ------I I I r----- /" 50 I 1 METRES The inner enclosure of the temple of Osiris at Abydos, excavated by Petrie. [Illust: H. Jaeschke, after Petrie, op. cit. [22], pl. XLIX] period,m measuring just over 170m square with a surrounding wall 3.50m thick (still standing in places 4.50 m high). In addition to this classic example, there are building remains of this reign at a number of important sites: Heliopolis yielded an alabaster offering table;21 and Abydos a foundation deposit and lintel22 to name but two. Another Ka House of Pepi 11 certainly existed at Zawiyet el-Amwat, capital of the XVIth, or Oryx, nome, where it is mentioned in the tombs of the 6th Dynasty officials - number 9, that of Shepseskau, having a specific reference. 23 Significantly, this is one of the towns mentioned in the lists of Sneferu cited previously. Middle Kingdom Climax Under the 11th Dynasty the construction of Ka Houses reached its apogee. The reunifier of Egypt Nebhepetre Mentuhotep was a very active temple builder and establisher of Ka Houses, and a whole series of these has been found in Upper Egypt.24 Starting with Abydos and working southwards, there are remains of at least nine temples erected by this king. In the Osiris temenos at Abydos Petrie found two granite altars set at 1.5 to 1.8m apart,25 as well as several inscribed blocks, suggesting a new temple was built here possibly with a stone sanctuary. Petrie therefore stated that all these sandstone blocks constituted `the first stone temple at Abydos'.% These blocks were discovered at points A and B on the plan, but cannot give more guidance than to show what may be the general area of the new temple (possibly north of the Old Kingdom brick temple if not actually on the site of the latter). In addition to these blocks, Habachi added some others which Lefebvre sent to the Cairo Museum in 1914." These consist of two further limestone blocks, found in the same place as those discovered by Petrie, showing Nebhepetre performing religious rites. There are thus seven sandstone and two limestone blocks of this king in all. Going on practice observed at other sites, the sandstone blocks could well have come from doorways of a shrine or even a gateway in a brick enclosure wall, these being constructionally better suited for this purpose than limestone. Those of limestone would, however, make the best material for actual wall surfaces carved with scenes to decorate halls and chapels. A most important clue as to the function of this quite sizeable temple is given in the lowest scene and inscription, depicted on Petrie's plate xxiv, which recounts part of the long lists of offerings of funerary type from a shrine in the building, ending with the words `for the statue (twt) of the king of Upper and Lower Egypt Neb- hepetre'. Here then, as elsewhere in Egypt, there is evidence for an important royal cult statue. What makes the dedication even more interesting is the reference on these blocks to the temple (hwt nir) of Khenti Imentiu, as it implies a structure at least partly devoted to the worship of the local god as well as the cult of the king himself during his lifetime, and certainly after death in the form of funerary rites such as one would expect at Abydos. Finally the limited number of blocks dating to Senuseret I at this site may be put down to chance but might imply that the 11th Dynasty temple was perhaps larger than that of the 12th. A similar situation existed at Dendera where a limestone chapel was discovered measuring Text from the Abydos blocks of Mentuhotep. [Illust: after Habachi, op. cit. [24], p. 181 82 JACF VOL. 5 2.20m long x 2.45m wide, with its axis running east to west. Habachi gives slightly different figures, but it is clear that this shrine is part of a larger building. The internal dimen- sions are 1.32m wide x 1.80m long, the greatest height 2.37m being at the front. That it stood for a long time is shown by the added inscrip- tions of Merenptah on the sides of the entrance, it being thus accessible at least eight centuries after its erec- tion. The original inscriptions on the entrance jambs are carved in incised hieroglyphs and are very important for defining its function. That to the right states, `the beloved of [Hathor, Lady of] Dendera, son of Re, Men- tuhotep, living like Re forever, the living god, foremost of the kings. The Ka House (hwt k3) of Mentu- hotep, which he made as his monu- ment for his statue (twt) "[Beloved- ofl-Horus" [living foreverl'. This makes it clear that here was aVKa House temple with at least one statue of the king in it, exactly as at Abydos. The text of the left jamb also refers to this Ka House and statue. Among the scenes on the walls inside the chapel are two showing Mentuhotep enthroned, one of which has offerings before him, as would of course be presented to his cult statue. In this context one must add an alabaster offering table of Nebhepetre, measuring 42cm long (52cm originally?), 34cm wide and 20cm thick, which was found, like the throne of a statuette of Pepi I, during work involved in the study of the blocks of the Roman mammisi.2s Remains of temple blocks reused in the South Palace at Ballas may indicate that Nebhepetre built a further monument here,29 but, if so, there is no certain evidence of its original location. Karnak must certainly have contained at least one temple of Mentuhotep, probably a major one, although evidence is at present equivocal as regards both its form and location. A massive granite offering table in the central store at the site is evidence for an important cult sanctuary.3o The table measures no less than 4.50m long, 1.75m wide and 75cm high, and could thus have even served a whole row of shrines if placed in front of them. A smaller offering table (Cairo 23007) and a large granite door lintel, reused in a wall marked J and K on Porter and Moss's plan, also attest to the king's building activity. The latter shows Nebhepetre seated on a throne, while near the Seventh Pylon, a further fragment was found bearing the cartouche of the king.31 Yet another interesting and mysterious monu- JACF VOL. 5 Right ment is a statue of Nebhepetre found by Legrain beneath the base of the flag mast in front of the southern face of the Seventh Pylon.32 The inscrip- tion on its stomach states that king Sebekhotep 111 `made it as his monu- ment to his father the king of Upper and Lower Egypt Nebhepetre, justi- fied, making a renewing of` the work of his predecessors Senuseret I1 and III. This statue, including base, is 1.95m high, wears the white crown and either bandages or a Sed-festival robe, but has its hands free, and recalls the rough work of some of the earlier Deir el-Bahri statues of the king. Here then is a later founded, if not contemporary, statue cult of the great king at Karnak. Finally, a lime- stone block fragment of 1 lth Dynasty work forming the edge of a cornice was found by Chevrier in the Third Pylon. 33 Two out of the other three Montu cult centres in the Theban province also have remains dated to this reign. A limestone door lintel from Armant, published by Percy Newbeny,34 has a balancing inscription reading: `Live! the king of Upper and Lower Egypt, Nebhepetre, son of Re, Mentuhotep, given life'. While not giving definite evidence as to the nature of the temple here, one so closely associated with the kiig's personal god would almost certainly have con- tained a Ka House chapel and statues of the monarch - like that at Dendera. Much more archaeological evidence exists for Tod where remains of two 11th Dynasty temples have been found in the platform of the Senuseret I temple.35 Fragments of at least seventeen or eighteen blocks and architectural elements of sandstone and limestone, as well as a pink granite statue base, were recovered by the excavators. These varied remains comprise in sandstone, por- tions of a colonnade with architrave, three doors and two lintels; and in limestone, wall reliefs decorated with incised carving, suggesting perhaps the exterior faces of chapels, as at Dendera. Bisson de le Roque suggested that these came from a temple mainly built of mud brick, as did Lintel of the Mentuhotep doorway at Armant. [Illust: after P. E. Newberry, op. cit. [34], p1.1(2)] 83 those belonging to his successor, but it is clear that considerable parts of the structure were stone, as in this king's Abydos Ka House temple. The structure would therefore have contained one or more limestone chapels preceded by a hall with at least four octagonal columns now surviving to a height of 1.63m but clearly originally loftier. These recall those used in the vast hall of the king's temple at Deir el-Bahri. The archaeologists noted the smaller scale of the figures shown in the scenes as compared with those of Sankhkare. In the limestone reliefs these average 52cm or one cubit high without head- dresses, while those of the later king are 64cm measured similarly, thus proving they belong to an entirely different structure. The columns taper and bear a text refemng to the king as `beloved of Montu, lord of Djerty (Tod)'. Among the lime- stone elements are parts of two or three doors (the larger one might be for the hall and the smaller for a chapel entrance). The statue base, measuring 1.24m long, 53 cm wide and 88 cm high, has been identified by Arnold as that of a small sphinx, but, if it was for a statue, then a royal cult place must be considered likely. Arnold has more recently iven a hypothetical reconstruction of this temple 3F which he suggests was mainly of brick for the exterior parts, and covered an area of 20m x 15m square. This includes its thick outer brick walls, the stone parts being considerably smaller. This general form has also been adopted by C. Desroches Noblecourt and C. Leblanc in a further study of the site. 37 Yet another temple of Mentuhotep existed at Gebelein where three blocks were found reused in the Ptolemaic houses, all now being in Cairo Museum.38 One has a scene in raised relief showing the king before an offering table bearing vases and collars with, perhaps significantly, his Ka spirit following him. He is called `Mentuhotep son of Hathor, Lady of Iunit (Dendera)'. It need hardly be added that this type of scene is anal- c. lOOm x 57.50m i 134mx78m 76m x 66m ogous to that previously described in the Dendera Ka-statue chapel. On the opposite bank further remains have been found at el-Kab. Within the late temple, the archaeological expedition found an almost com- plete door post and the lower part of another.39 In 1956 the remaining part of the latter was discovered. Blocks were found under the paving of the small temple and in the foundations of the 18th Dynasty temple naming Mentuhotep (with double cartouches). The text, in incised hiero- glyphs, reads: `... beloved of (Nekhbet) lady of Nekheb, son of Re, Mentuhotep, given life for- ever'. At the southern end of Egypt, Nebhepetre undertook considerable work on the island of Elephantine at Aswan. Earlier excavations, before the major clearance of the German expedition, had already shown that this must have been a large building or complex. The French found two huge sandstone door jambs, 5 m high, which, from their scale, probably came from an outer enclosure or temenos wall, presumably by its size built of brick (and recallin that of Senuseret 111 at Medamud cited later).4 In all, six sandstone blocks were discovered which fitted together to form a scene showing the goddess Satet offering life to the king who has Montu behind him wearing a solar disk. Each post also has a long vertical inscription in incised hieroglyphs giving the king's Horus name and calling him Nebhepetre `beloved of Satet, Lady of Elephantine'. Parts of what were ap- parently two octagonal sandstone columns were also found and are now in Cairo Museum, numbers 4-12-14-12. In 1946 Habachi found further elements of the same temple. An almost complete limestone door jamb41 and the lower part of another that almost certainly came from the entrance to a shrine chapel. The inscriptions, again in incised relief hieroglyphs, refer to Nebhepetre `beloved of Satet, Lady of Elephantine'. f 89m x 63m 40m+ x 30m+ xc*115m I Royal Builder I Location I Outer Wall excaiators to either the late Old Kingdom or Middle King- dom complex of buildings. 42 The eeneral area measures 1. Peribsen 2. Khasekhemui 3. Khasekhemui 4. Teti 5. Pepi I 6. Pepi I1 7. Nebhepetre (or Pepi 11) 8. Nebhepetre (or Pepi 11) 9. Sankhkare 10. Senuseret I11 Abydos Abydos Hierakonpolis Bubastis Bubastis Dakhla Oasis Abydos Elephantine Qma Medamud 109m x 99m Table I : Ka House enclosure dimensions. 84 During the more recent German excavations the area of the Satet temple is shown to have had massive brick walls surrounding an early enclosure (marked A and B on 112m x 56m the plan) and dated by the about u33m E-W x 25m N-S as extant, while to include the whole Satet temple in it this would have to stretch at least 40m N-S. The surrounding JACF VOL. 5 97m x 61m wall was 3m thick. The Middle Kingdom temple pro er had brick buildings and a stone f~undation,~? covering a general area of around 22 or 23m E-W x 16m N-S, with a stone floor about 6m square. Besides six stone elements belonging to Nebhepetre, there are seven of Antef 11, two of Antef 111, one of Sankhkare and four others of uncertain origin. The 11th Dynasty temple, or temples, here must con- sequently have had a long duration. The height of the column shafts was ascertained to be 2.65m.44 Comparative plans of the Satet temple's growth show that the 11th Dynasty period was compar- able in scale to that of the next dynasty.45 Indeed the final reconstruction plan of the temple shows that the new building of Nebhepetre46 measures nearly 15 m x 12.50 m - comparable dimensions to that of Tod. It had two doors in front, one leading to a square court or hall, with, at the rear, two niches flanking a statue room or sanctuary in the centre. The other door on the right side led to an oblong court with a peristyle of 5 x 3 columns, the whole being actually larger in area than the later building of Senuseret I of the 12th Dynasty. The walls of Mentuhotep's temple are just over lm thick, possibly 2 cubits, except for the sanc- tuary area which is embedded in a solid mass of material. To sum up, the Ka House buildings and associated temple structures of Mentuhotep form a so far unique picture of what these structures were like, and also constitute an unparalleled scale of such cult buildings - although those of the Old Kingdom rulers cited above may have been of the same order of magnitude and universal distribu- tion. Of the nine temples described here, two are certainly Ka Houses from their textual descriptions and two more probably so. The scale is varied, but that at Elephantine is likely to have been much greater than apparent if a probable shrine for Khnum also existed. These nine are evidence for only the area of the eight southern nomes of Upper Egypt, suggesting by average a probable thirty-six at least or even forty for the whole of Egypt. Again, if the four missing nome capitals in the area are included (which must surely have also had buildings of Nebhepetre), then the total is thirteen for eight provinces giving a figure of up to sixty for the whole of Egypt. The successor of Nebhepetre, king Sankhkare Mentuhotep, although he reigned only 12 years as opposed to the 51 years of his predecessor, was still able to build a large number of temple buildings. Indeed, too much can be made of the apparent shortness of his rule. The country being peaceful and unified, much work could be achieved in a period which is of similar length to JACF VOL. 5 the accepted duration of the reign of Seti I, when very many major works were erected if not completely finished. At Abydos, Sankhkare built another fine structure, the sand foundation that Petrie linked with it being now rejected by Kemp as unlikely to have belonged to this period (as shown by a reappraisal of the archaeolo ical bounded across thin walling by a square denoted DEFG) might, however, indicate its general site position. If so, it was about 27m E-W x 25m N-S. This would be an enclosure of similar size to that erected by Sankhkare at Qurna, or if only the temple sanctuary building then somewhat larger than that at Tod. Three blocks of this king were certainly reused on the site here, but most important is the fact that, while the pit on the west of this area marked `Deposit' on the plan was found to be empty, there lay near it the small alabaster foundation deposit tablet referring to the `Ka House (hwt k3) of Sankhkare'.49 Therefore, as Petrie stated, Sankhkare certainly built a new temple here, not just renewing or adding to a previous structure; and it equally belongs to the class of sanctuary under discussion in this article. A limestone block of Sankhkare50 has a scene, in incised relief, of the king offering before an altar with a goddess behind him. Another has a representation showing Osiris and a king wearing a feathered crown placed before many offerings, again carved in incised relief. This, although bearing a later king's prenomen (possibly that of Senuseret 111), Petie considered to have been altered from an earlier name - perhaps that of Sankhkare.51 A fragment gives the king's names and calls him `beloved of Wepwawet'.52 In all, four limestone blocks thus belong to the shrine of this king which is fewer than those of his predecessor. Quma is the most spectacularly sited temple in ancient Egypt being built at 365.8m above the Nile valley and is therefore in some way atypical. It is still the only completely known and recorded 11th Dynasty temple plan other than the Theban funerary examples.53 The rock was first scarped evidence). 47 The area shown in Petrie's plan 4i (as The Qurna temple of Sankhkare. [Illust: H. Jaeschke, after Petrie, op. cit. 1.531, pl. VI] N I 0 50 METRES 85 Looking north over the Valley of the Kings with the ruined stone huts of the Way Station beneath Meretseger (el-Qurn) in the foreground and the Sankhkare temple on the high ridge in the far distance (indicated by the small circle). [Photo: E. Rohl] and built up with boulders in the front area to make a level platform about 35.5m across N-S with a sloping entrance made in the centre of the east side. This leads through the east wall which was 24.74m E-W and twice as thick as those on the other sides of the enclosure - thus forming a pylon type front about 3.4m deep, possibly 6.5 cubits. The north and south walls measured about 1.6m (perhaps 3 cubits thick), and the rear wall slightly more at around 1.8m (say 3.5 cubits). So it follows that the enclosure within was probably intended to be 37.5 cubits (19.5m) E-W x 35 cubits (18.35 m) N-S. The temple shrine proper stood in the centre of the rear half and consisted of a brick built structure exactly 9.9m square. The front or east wall is again thicker then the other three - 1.8m (3.5 cubits) instead of 1.6m (3 cubits). The interior, built over a stone substructure, is divided into a front hall and three rear chapels, the central one being slightly wider (at 3.5 cubits) than the two flanking cells which were 3 cubits broad. Presumably the hall, at a lower level, had either two or four columns. The remains found by Petrie in this chapel are extraordinarily interesting and seem to have been the subject of a misrepresentation in his explana- tion. He claimed to have found a Heb Sed ritual object in the form of a dummy sarcophagus, with a separate lid, and `a seated figure, half life size' of the king. This last item significantly shows `the knee with the left hand resting on it, and a close fitting tunic coming down over the wrist'. Hence, Petrie suggests that this was a Sed-festival statue 86 wearing the distinctive robe used in the ceremonies.54 In fact, bands of inscribed text, cut in incised hiero- glyphs on the `sarcophagus', recall the side of a small shrine of the type which might have contained a statue like that found by Petrie. A simillar argument can be made for the remains of the attached toros moulding and cornice from the same so-called sarcophagus. If so, then here is a parallel with Mentu- hotep's shrine at Dendera. The remaining right hand jamb inscrip- tion is shown to be about 60cm long in the drawing, which would indicate a somewhat higher door post - just right for a smaller statue naos. The toros shown only on the left hand jamb again indicates that this would have been the southern or left hand shrine contained within the upper level brick chapel niche. The left hand inscription states, `[he made it as] his monument fo; Horus, he makes for him a given life [like] Re ...'; the right side may also mention `Hathor, Lady [of Den- dera]'. Associated with this lofty positioned shrine was a brick building that Petrie suggested was intended for use by the priests.55 It lay about one hundred yards behind the chapel, measuring 22.9 m long by 13.7 m wide, and containing three halls. Leaving aside a possibly similar temple found below the 18th Dynasty temple at Medinet Habu, there are a number of blocks from a temple or chapel of Sankhkare at Armant.56 These have scenes in raised relief suggesting they belong to the interior of chapels or a hall. However, the dedications are to gods like Montu of Hermonthis, Iunit, Nekhbet, Wadjit and Ptah, and do not contain specific references to the king's Ka House or a statue cult. Similarly, those elements discovered in the Senuseret I temple platform at Tod, like those of Nebhepetre, imply an important shrine of the period built here, but cannot as yet be linked directly with a royal Ka House, although this form of cult place is likely to have been used here.57 In all about thirty four pieces of masonry remain that belong to this building, for which Arnold has suggested another reconstructed plan, 58 measuring 30m x 15m, which must have included a front court (as the actual remains as restored cover an area in the plan of only 9.50m x 7.50m). The plan shows a hall and three chapels rather than a single shrine as in the Nebhepetre temple, and equates naturally with the Qurna sanctuary. Given JACF VOL. 5 that this temple could have had outer walls of stone and rubble or walls approximately the same thickness of those at Qurna, it need not have measured more than 10.5 m or 20 cubits square. The dedication on block number 1 160 indicates that we have here a temple for Montu, but the text on block 2125 states of the king that `he is fore- most among all living spirits (`kas'), appearing on the throne of Horus, given life forever' and hints of other purposes served by this complex. At Aswan Sankhkare, like his predecessor, built on Elephantine.59 A limestone block found in 1908 bears a scene in raised relief from a temple or chapel of this king. With a star border at the top, it shows him dedicating offerings to the deity. The Satet temple has therefore yielded evidence of his extending work there - if not actually building afresh. One last example of a Ka House is possibly to be found under the 12th Dynasty. At Medamud Senuseret 111 rebuilt the Montu temple and its enclosure using a design identical to the estab- lished Ka House layout. 6o Its inner enclosure wall measures about 97 m N-S x 61 m E-W, the walls being around 5.23 m or 10 cubits thick at the base. Round this stretched an outer enclosure wall about 109m N-S x 99m E-W over all, but much thin- ner, being only about 1.5m thick. This suggests that an enclosed space of around 105m or 200 cubits was intended. A restored gate belonging to the inner wall and now in Cairo Museum has, in section, a depth of 5.23m at the base, tapering to about 4.9m deep at a height of 3m from the ground. The height to the top of the door lintel measures around 3.65 m so that if the wall reached a height of 10.50m or 20 cubits at the top (double its base width) then the summit was still just over 4m or 8 cubits thick. The platform of the lime- stone temple is shown on the plan as about 31.5 m or 60 cubits N-S x 24m E-W, being thus a larger building than those described above, even allow- ing for the papyrus columns (coming from a court- yard) covering part of the area. While the inscrip- tions definitely link the dedication to the local god Montu, the numerous statues of the king found here might suggest a personal cult attached to this temple also. Nevertheless it would seem that such old foundations were now going out of favour and more temples became linked directly and solely with the worship of the major deities. In conclusion certain factors can be estab- lished in regard to the functions and existence of Ka Houses. They probably first appear in the Early Dynastic period (under the 2nd if not the 1st Dynasty) as an extension of the king's personal cult and funerary rites, distributed throughout Egypt. That this is an indigenous Nilotic and thus African feature seems apparent, and there is as yet JACF VOL. 5 no example of such a thing in Western Asia or elsewhere. They seem to be a possible insurance of the king's cults being able to continue should the shrine associated with the actual tomb be damaged or destroyed. Many such foundations are indicated by the 4th Dynasty Ka House lists, and during the 6th Dynasty a number of examples illustrate the form and design of these foundations. The scale and complexity of such buildings seems to have reached a peak during the long reigns of Pepi II and Nebhepetre Mentuhotep. It is under the latter that the immense extent and implications of this phenomenon can best be appreciated. With the advent of the 12th Dynasty, it seems that the Ka House system began to decline and eventually die away, due no doubt to different needs and the emphasis which began to be placed on what con- stituted gods' temples. However, even as late as the reign of Ramesses 111, as shown in my previous article in JACF 4,63 there still existed cult statues in all the king's temple foundations throughout the empire, as far as Beth Shan in Palestine to the north and down in Nubia. 0 Notes and References 1. 3: E. Ayrton er al.: Abydm IU, pp. 4-5, pl. VII. I. E.'&belL &r:erakonpolis I (London. 1900); J. E. Quibell and F. W. Green: Hierakonpolis II (London. 19oL). plan pl. IXXIV. pp. 19-20.25: M. J. Garstang: `Excavations at Hierakon lis, at Esna, and in Nubia' in MAE 8 (1907). pp. 136-37, pls. 7-Q M. I. Garstang: `Excavations at Hierakonplis Fort. 1905, a Preliminaly potel in BMMA section 11, Nov. 1935, 40-4; B. I. Kern . Excavations at Hierakonpolis Fort, 1905: A %hinary Note' in J& Garstang, op. cd. [3] (Nov. 1935). Ibid.. p. 11, pls. XL, XU. 2 Ibid. 2-3, IS. V-VI. 49 (1963), pp. 24-8, pls. N-V. 4. 5. 6. Quibell: op. ca. 131, p. 11. pi. XXXM. ---_I I I 50 - METRES Reconstructed plan of the temple of Senuseret 111 at Medamud. [Must: D. Rohl, after Robichon & Varille, op. cit. 1601, fig. 3J 87 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. A. Fakhry: The Monuments of Sneferu at Dashur vol. II. The Valley Temple: Part I (Cairo, 1959). lists, pp. 17-58. A. El-Sawi: Excavations at Tell Basta (Prague, 1979). pp. 75-6. Ibid., see figs. 164-65. L. Habachi: Tell Basta (Cairo, 1957). pp. 11-32. Loc. cit. Habachi. op. cif. [lo]. fig. 2. G. Jequier: L.e Monument Fmerawe de Pepi II, vol. JII (Cairo, 1940). see pls. 19 (south wall), 21 (south wall), 22 (south wall), 23 (south wall), 24 (north wall). See pls. 25-6 (north wall). P. Posener-Kritger: 'Les travaux de l'hstitut Fragais d'Archhlogie Orientale en 1985-1986' in EIFAO 86 (1986). pp. 371-72; EIFAO 87 L. Pantalacci: 'Un dtcret de Pe i 11 en faveur des gouvemeurs de l'oasis de Dakhla' in EIFAO 8?(1985), pp. 245-55. Ibid., see plan fig. 1. Ibid., fig. 2. L. L. Giddy: Egyptian Oases (Warminster. 1987). p. 184 plan 2; see also I. Vercouter: 'Les travaux de l'Institut franpis d'Archhlogie Orientale en 1977-1978' in EIFAO 78 (1978). pp. 575-76; EIFAO 79 189-205; and L. L. Giddy: 'Rapport pklhmalre sur la quatrieme campaign de fouilles B Xyn Asfl (Oasis de Dakhleh)' in ASAE 69 (1983). pp. 103-12, fig. 4. W. von Bissing: Steingefasse (Cairo Cat.), p. 114. W. M. F. Petrie: Abydos 11, (London, 1903). p. 12, pl. XXI, p. 31, (1987). pp. 300-1; EIFAO 88 (1988). pp. 183-84. (1979). pp. 31-9; EIFAO 80 (1980). pp. 257-69; SIFAO 81 (1981). pp. dS. XM-XX. 23. b I, 57(9), Text II, 63. Habachi: op. cif. [lo], pp. 39-40. 24. See especially L. Habachi: 'King Nebhepetre Mentuhotep, His Monuments, Place in History, Deification and Unusual Representation in the Form of Gods' in MDAIK 19 (1963). p. 16-52. 25. Petrie, op. cif. [22], p. 12 and plan on pl. JV. 26. Ibid. see pls. XXIV, XXV top sections. 27. Ibid. p. 17. 28. G. Daressy: 'Chapelle de Mentouhotep III d Dendtra' in ASAE 17 (1917). pp.226-36, pls. 1-3, Habachi, op. cit. [24], pp. 19-28. pls. IV-VIII, Porter & Moss VI, plan p. 40 for location of chapel; H. G. Fischer: Dendera in the Third Millennium E.C. down to the Theban Domination of Upper Egypt, (New York, 1968). pp. 52 note 209, 126, & 186. 29. Habachi, op. cif. [24], pp. 28-30. 30. Habachi, op. cif. [24]. 31. G. Legrain: 'Notes d'inspection' in ASAE 5 (1905). p. 134-36, fig. 1. 32 G.Legrain: 'Notes d' inspection' in ASAE 7 (19077, 33. H. Chtvrier: 'Rapports sur les travaux de Kamak' in &?6$947), 34. !. E. Newbeny: 'Extracts from my Notebooks. VII' in PSEA 25 (1903). p. 362, pl. l(2). 35. F. Bisson de la Roque: Tod 1934 a' 36 (Cairo, 1937). py 62-78, see numbers 21 16. 1168 (fig. 16). 1169 (fig. lo), 1126 (fig. 1 ), 1170 (fig. 1510. 36. D. Amold: 'Bemerkungen zu den friihen Tempeln von El-Tijd' in MDAIK 31 (1975). pp. 175-86, Abb. 1. section Abb. 2. 37. C. Desroches Noblecourt & C. Leblanc: 'Considtratim sur l'existence des divers temples de Monthou d travers les Iges, dans le site de T6d. Etat de la question en Octobre 1983' in EIFAO 84 (1984), pp. 81-7, Kab and Gebelen' in PSEA 15 (1983). pp. 497-98; . 150, pl. XXXIV. 18). 1171 (fig. 23). 2120, 1561-2, 2117,2121, 1543,2128, 1544, 1542, 1. XXXIII for, lan. 38. 8. W. Frazer: 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. G. Daressy: 'Notes et remarques' in Rec. Trav. 14 (1893). p. 26 & Rec. Trav. 16 (1894). p. 42, Habachi, op. cif. [24] figs. 15-17, pls. Xb-XIa-b, Porter & Moss V, p. 163. I. Capart et al.: Fouilles d'EI Kab (Bruxelles, 1940). pp. 22-3, pl. 30a-b. H. Gauthier: 'Remarques sur la XIo Dynastie' in BIFAO 9 (1911). p. 123; Habachi. op. cif. [24], pl. XIIa-b. Cairo Mus. number 41557. fig. 19 and pl. XIII. Gauthier, op. cif. !40], fig. 18a. W. Kaiser et al.: Stadt und Tempel von Elephantine' in MDAIK 28 (1973). p. 163, plan Abb. 1. W. Kaiser et al.: 'Stadt und Tempel von Elephantine' in MDAIK 31 W. Kaiser et al.: Stadt und Tempel von Elephantine' in MDAIK 32 W. Kaiser et al.: 'Stadt und Tempel von Elephantine' in MDAIK 33 (1977). m. 69-71. Abb. 2. (197.5). pp. 45-9, bbb. 1. (1976). pp. 72-3. ,-- I. rc- -- W. Kaiser et al.:-'Stadt id Tempel von Elephantine' in MDAIK 44 (1988). pp. 152-3. lan Abb. 6, for Senusret I building Abb. 7. B. J. Kemp: 'The &iris Temple at Abydos' in MDAIK 23 (1%8), p. 138-55, see es cially . i40 and pl. XLmA-B. Fetrie, op. cit. [225Iep1. LJ: Ibid. p. 33, pl. XXII.5. Ibid, p. 33, pls. XxIII.2 and XXV bottom. Ibid, pl. XXIII.3. Ibid. pl. XXV middle; see also Habachi, op. cif. [24]. W. M. F. Petrie: Qurneh, (London, 1909). pp. 4-6, plan pl. VI also views 1. V. Ibid. pf W. 55. Ibid, ilan pl. VIE. 56. R. L. Mond & 0. H. Myers: Temples of Armant, (London, 1940). pp. 2. 22-3. & 49-50. Dls. XCIV. XCV. XCVI. XCW. Xo(. 57. Bisson de la Roque, op. cif. [35], pp: 79-1od, fig. 38, 41. 46. 48. 51, 58. Lisson de la Roque, op. cif. 1351, Abb. 3. 59. J. ClCdat: 'Un nouveau monument du roi Sankhked' in Rec. Trav. 31 (1909). p. 65; Kaiser et al., op. cit., [44], pp. 73-4. 60. C. Robichon & A. Variqe: 'Mtdamoud. Fouilles du Musk du Louvre, 1938' in Chronique d'Egypte 27 (1939). pp. 84-5. plans figs. 2-3. 61. F. Bisson de la R ue Rapport sur les Fouilles de Mkdamoud 1929, (Cairo, 1930). plsb-XIII. 62. F. Bisson de la Roque: Rapport sur les Fouilles de M&hmud 1928. (Cairo, 1929), pl. E, and Rapport sur Ies Fouilles de Mkdomoud 1930, (Cairo, 1931). pi. N. 63. E. P. Uphill: 'Papyrus Hams: A Funeraty Endowment?' in JACF 4 Addendum: Since this article was written, a paper by David O'Connor has appeared which discusses the function of sane of these Ka temples, but using a rather different approach. He disputes B. Kemp's theory relating to early temples in a 'preformal mode' being typical at some provincial sites. Nevertheless, he accepts the temples cited above at Abydos. Bubastis and Heirakonpolis but, hke Brovarski. makes what may be an arbitrary distinction between such royal ka-chapels and the local god's temples, by postulating separate structures for each, even though the evidence does not seem to support more than one building as a rule on a particular site. See 'The Status of Early Egyptian Temples: An Altematlve Theory' in R. Friedman & B. Adams (eds.): The Followers of Horw, Oxbow Monograph 1. xxc. (1991). pp. 41-8. 20 (1992). pp. 82-98. The desolate ruin of the Shunet el-Zebib at Abydos. [Photo: D. Rohl] 88 JACF VOL. 5