SUDAN & NUBIA

1

2

SUDAN & NUBIA

1

2

SUDAN & NUBIA

9

Excavation of the LateKushite and Medievalsettlement on Umm MuriJohn Payne

The site, designated 3-J-5, is located on the uninhabitedisland of Umm Muri (Colour plate IV) and was first dis-covered during the 1999 SARS survey (Welsby 2003, 5-7,pl. 2.1). It is defined by a large scatter of ceramics, stonecereal grinding bases (Usai 2003, 106-9, pls 5.10, 5.11) anda number of mud-brick and dry-stone walls.

The initial survey was followed in 2003-2004 by a smallevaluation of the site, consisting of three test pits exca-vated in order to assess the depth and nature of the ar-chaeological remains (Fuller 2004, 8-10). This evaluationindicated that the archaeological deposits across the wholeof the site were likely to be quite shallow and comprised ofmud-brick and stone structures of late Meroitic date. Withintest pit 7HC, these mud-brick structures were seen to over-lay a series of pits and post-holes which althoughstratigraphically earlier than the mud-brick structure wereotherwise undated.

The aims of the 2004-2005 season were to reveal a largerarea of the site in plan by linking the three previous testpits, selectively excavate areas within this larger area in or-der to better understand phasing and construction techniqueswithin the late Meroitic settlement, attempt to date and bet-ter understand the post alignments and ascertain whetherthe large dry-stone wall visible at the eastern extent of thevisible remains did indeed mark the eastern limit of thesettlement, as suggested by the previous evaluation. Along-side this work it was decided to produce a basic 1:200 scaleplan on which was plotted the limits of the ceramic scattersand which served to highlight areas of bare rock within thesettlement area in which the archaeological horizon had beenlost.

The work was undertaken by a team of eight archaeolo-gists and an NCAM inspector1 plus a local workforce ofup to 40 persons over a period of around ten weeks. Withinthis period, an area of approximately 1800m2 was exca-vated which probably represents approximately a 15% sam-ple of the archaeological remains of the Umm Muri settle-ment (Colour plate V).

Phase 1The earliest phase recognised was located within the centralarea of the excavation directly to the north of evaluationtest pit 7HC. It consisted of a suite of three roughly squarerooms each measuring approximately 3 x 3m with distinc-tive rounded corners (Plate 1, Figure 1). The walls, con-structed of small stones set into a mass of mud, survivedto a height of around 300mm and were rendered bothinternally and externally with a fine mud plaster. The plasterwithin the interior of the structure showed evidence of sim-ple painted decoration. Floor surfaces were also revealed,although no internal features were noted. It seems likelythat the rooms revealed do not represent the original limitsof this structure, and it is possible that it extended bothnorthwards and westwards, although due to time constraintsthis could not be proven.

Directly to the south of this structure were the post-holes and pits partly revealed in the previous season’s evalu-ation. Within these, curvilinear alignments of post-holes couldbe recognised and some appeared to abut the southern wallof the Phase 1 building (Plate 2). It is therefore likely thatsome of the post-holes represent ancillary structures asso-ciated with this phase. One pit contained a large number ofloom weights (Plate 3).

Phase 2An expansion or reconstruction of the Phase 1 structure tothe west (Figure 1). This phase also marked a change inconstruction technique from stones set in mud to mud brick,which continued throughout the subsequent phases. Tworooms were recognised dating to this phase; both were rec-tangular in plan and aligned north-south. The easternmostof the two abutted the western wall of the Phase 1 struc-ture, and although clearly of a later phase there was someevidence to suggest that they was built directly onto an ear-lier foundation. Within the easternmost of these additionalrooms evidence of an internal division was recognised in

1 The team consisted of Nick Corcos (archaeologist), Kat Manning(archaeologist, animal bone specialist), Eric McCann (pottery special-ist), Paul Murray (archaeologist), Dave McNicol (archaeologist), JohnPayne (director), Adam Stone (archaeologist), Nialla Fayer-Kerr (ar-chaeologist, BIEA graduate scholar) and Abdel Rahman Abdel FatahAli Siddiq (NCAM inspector). The team was greatly assisted for partof the season by Nasreen Sedek Yahya. Neal Spencer joined the projectfor one month.

Plate 1. The Phase 1 building looking north.

10

SUDAN & NUBIA

11

the form of a small brick structure located in the north-west corner of the room. This consisted of an ‘L’-shapedtank which abutted and utilised the wall of the room as partof its construction. This tank was rendered both internallyand externally with a fine mud plaster. The function of thisfeature is uncertain, although its similarity to storage binsseen at Meinarti, for example, may give a possible explanation.

Phase 3This phase appeared to represent a dramatic expansion ofthe structure which involved the demolition and levellingof the Phase 1 building and the construction of a suite ofat least ten additional rooms (Figure 1). The Phase 2 build-ing appeared to be incorporated within this new phase ofconstruction. All of the new rooms were rectangular in plan

with the majority being aligned east-west. Seven of theserooms were to the east of the Phase 2 structure some ofwhich overlay the Phase 1 building. The three remainingrooms identified within this phase were located to the southof the Phase 2 building. Within two of these three roomsthere was clear evidence of in-situ burning which possiblyindicates areas associated with food preparation. If this iscorrect, one could hypothesize that these structures weredeliberately detached from the main dwelling as a safeguardagainst fire, a phenomenon still seen within the traditionalstyle buildings used today. Floor surfaces within this phaseof construction were sporadic, with later pits causing areasof truncation. The intervening space, between the threeisolated rooms and the associated structures, was linked bya wall which appeared to form part of a larger enclosedarea which, due to its size, has been interpreted as an opencourtyard.

Phase 4The structures identified within the Phase 4 constructioncould actually represent two or three additional phases as itincludes two separate expansions of the Phase 3 buildingboth to the west and south, plus the insertion of additionalwalls within the Phase 3 structure (Figure 1).

The western extension consisted of an ‘L’-shaped wallwhich abutted the western wall of the earlier Phase 3 struc-ture. This ‘L’-shaped wall represents the southern and west-ern walls of a new structure. The eastern side of this newbuilding would have been formed by the existing Phase 3building. The interior of this new structure and its northernextent would have lain on an area of higher ground to thenorthwest of the Phase 3 building. The higher location ofthis structure has resulted in its northern extent being totallydestroyed by wind erosion. A single fragment of abuttingwall gives some indication of the existence of the internaldivisions within this structure.

The southern expansion associated with this phase com-prised a suite of four rooms which were partly excavated aspart of the previous season’s work (test pit 7HC). Theserooms which are roughly square in plan, partly abut thePhase 3 building. The walls were particularly thick andalthough they survived to a height of around 400mm theyshow no evidence of door openings. This possibly couldindicate that originally they were accessed from above, andcould represent storage areas. This construction phase wasseen to overlie the post-hole structures which are thoughtto be associated with the Phase 1 building. Therefore withthe construction of this phase it is clear that the post-holestructures were no longer in use, although it is likely thatthey had disappeared earlier, contemporary with the demo-lition of the Phase 1 structures.

Phase 5Another westwards expansion to existing structures utilisedboth the western wall of the Phase 3 structure and the south-ern wall of the Phase 4 structure in its construction. It

Plate 2. Post-holes associated with the Phase 1 building.

Plate 3. Loom weights from a pit associated with the Phase 1 building.

12

comprised a suite of five rectangular rooms aligned bothnorth-south and east-west (Figure 1). The earlier walls ontowhich this construction phase abutted showed evidence ofexternal rendering which continued behind the later abut-ment. This clearly indicates a later addition rather than phas-ing within a continuous construction. The preservation ofinternal surfaces within this phase was sporadic, due to acombination of variable erosion and later pit cutting. Thisadditional suite of rooms possibly formed a distinctdwelling.

Phase 6This phase is likely to represent two distinct phases as thequality of construction between these additions differedgreatly. However, as no stratigraphic link existed betweenthem, this is at present impossible to prove. Both abuttedthe Phase 5 structure (Figure 1).

The first abutted its western wall and again comprised asuite of five rooms. As with the previous phase all wererectangular in plan and aligned both north-south and east-west, although the area covered is approximately half thatof the Phase 5 structure. The second of the previous sea-son’s test pits was located within this structure. Within twoof the rooms which comprise this suite, successive floordeposits were revealed plus evidence of entrances whichagain showed successive episodes of raising and resurfac-ing. This suggests that the structure was relatively long lived,although it is clearly late within the overall phasing of thesite.

The second suite of rooms was located to the south ofthe earlier phase building. This again comprised rectangularrooms aligned both north-south and east-west which abut-ted the earlier buildings. The construction of these struc-tures did appear to be far less robust that the surroundingbuildings, their wall being only a single brick length thick.Again floor surfaces were sporadic and later pit cuttingseemed particularly severe, making any more detailed inter-pretation difficult.

The eastern part of the siteThis area included the third of the test pits from the previ-ous evaluation which had focused on the large stone wall(Plate 4), which was thought to mark the eastern limit ofthe settlement. This wall, which was over 1m wide, sur-vived for a considerable distance both north-south and east-west. However the area excavated and recorded in detailmeasured approximately 35m north-south x 25m east-west.This excavation revealed the existence of mud-brick struc-tures on either side of the stone wall. To the east of thelarge stone wall a suite of at least three large rectangularrooms was investigated. Each measured around 7 x 5m andalthough they were constructed from a combination of bothmud brick and mud and stone they seem likely to representa single phase. These three rooms incorporated the stonewall as part of their construction. This indicated that thelimit of the settlement lay further to the east than was origi-

nally anticipated, although it is possible that the large stonewall did at some time represent the eastern limit of thesettlement. To the west of the stone wall further evidenceof rectangular mud-brick structures was revealed which againutilised the stone wall within their construction (Colour plateVI). However truncation within this area was again severewith the majority of the structures surviving for only a sin-gle brick thickness, and any floor surfaces surviving as iso-lated patches. In these rooms were a number of potteryvessels containing animal bone and charcoal and a sequenceof ovens (Plate 5).

As a consequence of this truncation very little phasingwas apparent within this area of the site; however, a seriesof pots set into pits which clearly cut through some of thewalls did seem to indicate the presence of later, althoughnow eroded, dwellings.

The area which linked this area of activity and the phased

Plate 4. The stone wall looking south from the junctionof the north and east walls.

Plate 5. A sequence of large cylindrical-bodiedpottery vessels re-used as ovens.

SUDAN & NUBIA

13

structures to the west was excavated but unfortunatelyrevealed only the scantest evidence of archaeologicalremains. Despite this it seems likely that this apparently blankarea did originally contain a similar number of structures,which have suffered from severe erosion.

Although the majority of the site appeared as a wellorganised, fairly regular settlement, a series of poorlyconstructed curvilinear walls were seen within the north-western limit of the excavation. These appear to intrudeinto the proposed alignments of the partly eroded regularstructures. It, therefore, seems probable that these irregularstructures represent the only recognisable remains of a laterphase of settlement which has elsewhere been totally erodedaway. The function of these irregular walls seems most likelyto be associated with animal management. The presence ofChristian period ceramics on the surface of the site suggesta date for this phase of occupation.

An initial interpretation of the evolution of the UmmMuri settlement would envisage an early phase of irregularmud and stone and timber-post structures which were gradu-ally superseded by regular mud brick and stone buildings.The stratigraphic evidence appears to suggest that thismore regular construction phase commenced in the north-eastern part of the island (the higher ground) and graduallyevolved both westwards and southwards (down slope).Whether this observation of higher ground being favouredfor early settlement can be applied to all of the unexcavatedareas of higher ground, which are located to the east of theexcavation, remains unclear. This phase of regular settle-ment structures is likely to have been replaced at some timein the Christian period by a more organic looking settle-ment of which only a small portion was recognised in theseexcavations.

Among the finds were a small number of painted Meroiticsherds (Colour plate VII) and what may be the lid of aglazed composition box with a reclining leopard (Colour plateVIII).2

Acknowledgements

The project would like to thank the British Institute in East-ern Africa for its financial support. I would also like to thankmy own employer (Oxford Archaeology) for its generousdonation of the majority of the site equipment used in theexcavation.

2 For what may been a very similar piece, also thought to be a box lid,from Kumbur, north of Aksha, and dated to the 1st century AD, seeMaystre 1975, 91, fig. 35; Wenig 1978, 256, 258; Wildung 1997, 363.

BibliographyFuller, D. Q 2004. ‘The Central Amri to Kirbekan Survey. A prelimi-

nary report on excavations and survey 2003-04’, Sudan & Nubia 8,4-10.

Maystre C. 1975. ‘Découvertes récentes (1969-1972) près d’Akasha’,in K. Michalowski (ed.), Nubia: Récentes Recherches, Warsaw, 88-92.

Usai, D. 2003. ‘The ground stone tools’, in Welsby 2003, 106-109.Welsby, D. A. 2003. Survey above the Fourth Nile Cataract. Sudan

Archaeological Research Society Publication no. 10. London.Wenig, S. 1978. Africa in Antiquity II. Brooklyn.Wildung, D. (ed.) 1997. Sudan. Ancient Kingdoms of the Nile. Paris -

New York.

2

SUDAN & NUBIA

1

2