Tasian vessel.  Thanks to Digital Petrie at Contents


As I have said elsewhere, one of the problems with providing a coherent summary of the occupation of the Eastern Desert, is that large areas of it have been neglected, for many reasons, by archaeological survey teams.  This means that there are massive gaps in knowledge about how, when and where the Desert was occupied.  As Bomann and Young comment after their survey of the Middle Eastern Desert wadi Abu Had (1994, p.44), “the Eastern Desert could be a very fruitful area for prehistoric resaerch.  That archaeologists have neglected its prehistory for so long is, ultimately, their loss.”

Sir W.M. Flnders Petrie effectively divided the ancient Egyptian past into three main sections when he defined the Predynastic, with a prehistoric period preceding it and the Dynastic or Pharaonic period following it.  His definition of a pre-literate “Predynastic” was done on the basis that the material of the Amratian (Naqada I) and Gerzean (Naqada II) displayed considerable social and cultural complexity, and therefore needed a term to characterize them.  I have therefore left this scheme in place, and on this page I am looking only at the pre-Naqadan phases of the Egyptian past, which includes the Palaeolithic, Neolithic, Tarifian, Tasian and Badarian periods.

I have divided this page into a number of chronological headings, as shown in the Contents.  To help clarify matters, I have put together a set of timelines which can be found in Appendix C – they might make matters somewhat less opaque.  As I explained in the introduction, I am using a terminology that is convenient for the layout of this section, and enables me to communicate the information in a structured way, but it does not always correspond precisely to other schemes used in other publications.  The table shown on the Archaeology Introduction page explains how the different schemes relate to each other. 

I am also using Vermeersch’s 2002 scheme for the Egyptian Nile Valley as a baseline against which all of the Early and Mid Holocene phases can be compared (from the Late Palaeolithic until the Badarian).  This scheme is shown both on this page and reproduced in Appendix C (see the excellent article by Vermeersch – 2002, p.27-40 – for full details).





Vermeersch Phase (2002)

Earlier Prehistory


Lower Palaeolithic

c.500,000 – 250,000 BP



Middle Palaeolithic

250,000 – 70,000 BP



Upper Palaeolithic

35,000 – 21,000 BP



Late Palaeolithic

21,000-13,000 BP

Phase 1
Late Palaeolithic Occupation


Wild Nile

13,000-12,000 BP

Phase 2
Wild Nile Phase


Occupation Hiatus

13/12,000- 8500 BP

Phase 3 & 4
Phases of Population Reduction and Early Desert Occupation



8500 – 7500 BP

Phase 5






Post-Epipalaeolithic occupation hiatus

7500-6900 BP

Phase 6
Phase of Minimal Occupation

Later Prehistory

Earlier Phase

Early Eastern Desert Neolithic

6900-5300 BP

Phases 7 & 8
Phases of Desert Reoccupation and Lower Egyptian Agricultural Reconversion


6,300 BP

Faiyum Neolithic

6500 – 5500 BP

Later Phase


c.5870-5860 BP

Upper Egyptian Reconversion


5500 – 4900 BP


I have added a few words of introduction to each of the sections below, to put them into a broader Egyptian context, but more detailed information can be found in a number of standards texts (e.g. Midant-Reynes 2002; Hoffman 1979; Wengrow 2006; Hendrickx and Vermeersch 2002) and on

Rock art of all periods is handled quite separately in a dedicated Rock Art section on this site, the reasons for which have been explained on the Home Page.


Early Prehistory

Lower Palaeolithic

The earliest period of archaeological time is the Lower Palaeolithic in Old World terms, which includes the Oldowan and Acheulean industries.  In some parts of Africa the Lower Palaeolithic goes back several hundred thousand years.  In Egypt, it probably extends from 500,000 to around 250,000 BP. 

The principal Lower Palaeolithic industry of ancient Egypt is the Acheulean.  Acheulean tools have been found throughout Egypt.  However, no detailed information is available about the origins of the groups, or their lifestyles, and no human remains have been found in Egyptian contexts of this date.  The Acheulean in Egypt is thought to have come to an end with the downturn in climatic conditions.  For more information about the Lower Palaeolithic in Egypt, see the Lower Palaeolithic page on

In the Wadi Abu Had survey of 1992, Bomann and Young (1994) identified two definite examples of handaxes made of flint.  One was very well preserved, from site WAH-9 and a more eroded example was found at WAH-13.  Both handaxes “indicate that their cutting edges were created by alternate flaking using soft hammer, direct percussion techniques” (1994, p.38).  A large flint scatter at WAH-8 included items of grey-brown flint with a “potential handaxe” possibly dating to the Lower Palaeoltihic, although the other finds seem to date to a later period (1994, p.40).

Although Acheulean remains are also known from the Red Sea Hills, including in the El Gouna area, I have not yet located any information that would permit a description of the distribution or composition of such assemblages.

Middle Palaeolithic c.250,000 – 70,000BP

The Middle Palaeolithic of Egypt is quite clearly a complex archaeological period, with a number of different periods, industries and subsistence strategies represented. It is far more varied than the less homogenous than the preceding Acheulean.  

There are a number of different chronological schemes for dividing the Middle Palaeolithic, but one of the most useful is that proposed by Hendrickx and Shaw 2000 (p.20):

  • Early Middle Palaeolithic 250,000-150,000BP (including the Nubian Middle Palaeolithic and Sangoan),
  • Mid Middle Palaeolithic 150,000-80,000BP (including Khormusan, Denticulate Mousterian, Egyptian Group K, Egyptian Group N, Nubian Mousterian and Saharan Mousterian)
  • Late Middle Palaeolithic 80,000-70,000BP (including Halfan and Safahan/Levallois Idfuan).

Vermeersch and Hendrickx (2000, p.21) describe the Middle Palaeolithic as a period which experienced intervals of hyperaridity and lacustrine episodes.  During more humid periods, when most of the occupation evidence is found, “there were permanent lakes in the Western Desert or, in some intervals, seasonal playas, fed by local rainfall of up to 500mm per annum . . . The area was abandoned during the periods of hyperaridity.”  The climatic episodes during the Middle Palaeolithic allowed for much wider occupation of the Sahara, and for movement of groups over large areas, allowing for the wide distribution of similar industries.  As outlined in the Environment section in the Eastern Desert, this included periods of rain storms and the establishment of permanent pools. 

The Middle Palaeolithic of Egypt is identifiable by a number of distinctive lithic industries, some of which employed new techniques for reducing stone pieces to tools.  Industries include the Halfan, the Edfuan, the Nubian N-Group and the Saharan Aterian in the Western Desert.  The Aterian never reached the Nile and would therefore be very unlikely to appear in the Eastern Desert (for a discussion of the controversial Aterian industry see Kleindienst 2001).  The most important of the Middle Palaeolithic tool reduction techniques is known as the Levalloisian.  For more information about the Middle Palaeolithic of Egypt see the Middle Palaeolithic page on

Some flint scatters represent occupation in various parts of the desert.  Site GSAH-3  in the Wadi Abu Had in the Middle Eastern Desert , described by Bomann and Young (1994, p.40), has produced a tortoise core and and a knife that appear to be Levalloisian (Middle Palaeolithic), and part of an industry which concentrated on the production of blade tools.  In the southwest to southeast orientated Wadi Deir Bolos (off which the Coptic Monastery of St Paul is located), Middle Palaeolithic finds indicate that “at least the northern section of the Eastern Desert must have been passable for groups of hunter-gatherers during Middle Palaeolithic times” (Dittman 1993, p.149).  Sites consist of undisturbed knapping places within old gravels, but the geomorphological data can only suggest that they belong to a period earlier than 30,000 BC (Dittman 1993, p.149). 

The Middle Palaeolithic is more richly represented in two main areas in the Eastern Desert, both in the middle region in the Red Sea Hills – the  Wadi Bili and the Wadi Sodmein.  The sparsity of the sites represented is most likely to be an artefact of the small number of archaeological surveys carried out in a very limited area, rather than a reflection of genuine occupation patterns.

Bili Cave

The Middle Palaeolithic is well represented in the Red Sea Hills at two sites in particular – Sodmein Cave and in the Bili Cave area (near El Gouna) on the Red Sea coast. 

In the Bili Cave area in Wadi Bili in the Red Sea hills a number of concentrations were found, and appear to include an Early and a less well represented Late phase.  A single artefact, possibly a Middle Palaeolithic notched flake, was found within the cave.  A second artefact, a large Levallois flake was found outside and below the cave accompanied by some other artefacts. To the south of the probably Neolithic site ME03/10/24  a single site of Middle Palaeolithic materials appears to belong to the earlier stages of the Middle Palaeolithic. 

In the area as a whole several concentrations of Middle Palaeolithic Levallois material were found, and Vermeersch et al (2005) suggest that these probably belongs to the earlier part of the Middle Palaeolithic.

Sodmein Cave

Sodmein Cave has produced the most exciting finds from this period in the area.  It is located 35km NNW of Quesir, and is the largest known cave in the Eocene limestone outcrops of the Gebel Um Hammad.  The cave was apparently created by solutional and mechanical widening of a karst gallery outlet in a cliff of Theban limestone.  It is located in the lower part of a cliff overlooking the lower part of the Wadi Sodmein.  The Gebel Umm Hammad hogback probably provided the solutional waters that formed the cave in the form of run-off.  Deposits completely fill the cave in places, and can reach up to 5m in depth.  Due to the dry conditions of the deposits, both floral and faunal remains have been well preserved and have been used to build the environmental sequence described elsewhere on this site.

Four Middle Palaeoltihic levels have been identified (Van Peer et al 1996). The oldest layer, Middle Palaeolithic Level 5 (MP5), consists of a fireplace of several phases containing burnt bone.  The bones provide a great insight into the diet of the occupants and includes

  • Syncerus caffer (buffalo)
  • Loxodonta Africana
  • Tragelaphus strepisceros (kudu)
  • Melivora capensis (honey badger)
  • Gazella dorcas
  • Procavia capensis (rock dassie)
  • Crocodylus sp.
  • Clarias sp. (catfish) pectoral spine
  • Melanoides tberculata shell

Plant remains include acacia, clematis, grasses, and Ficus sp.

Bifacial stone technology is represented by a fragment of foliated handaxe similar to Nubian Middle stone Age industries (N-Group).  A Nubian type 2 core with a central ridge created by transverse flaking of the upper core face provides a “stratigraphic argument for the greater age of this Levalloisian method relative to the Nubian 1 method” (Van Peer et al 1996, p.155). 

The second oldest layer of archaeological remains, Middle Palaeolithic Level 4 (MP4), also appears to represent an N-Group industry similar to that found in Nubia with the characteristic Nubian points and cores, as well a a classical Levallois flake with a band of red ochre all over the central flake area.  Level 4 is dated to c.44,500BP.  Next in the sequence are tools which show an emphasis on blade manufacture, with some Levallois components.  Tools from this layer include denticulates, burins and Emirreh points.  The Emirreh points are interesting because they have never been recorded in the Nile Valley Middle Palaeolithic, but instead suggest connections with southwest Asia.  Vermeersch et al (1994, p.38) suggest that the Middle Palaeolithic material from Sodmein indicates the possible use of the Red Sea coast as a “passageway from East Africa to southwest Asia”. 

Next, Level MP3 was very well defined, immediately overlying a date of 45,000BP (UtC-3317).  The lithic industry shows both classical Levallois reduction and the Nubian 1 method.  Blades are rare.  Truncated-faceted pieces and Levallois flakes suggest, again, an affiliation with N-Group groups of the Nile Valley.

The next level, Level MP2, is well represented at Sodmein.  Levallois technology is present in the form of one blade, retouched and probably tanged, appearing similar to Aterian elements at Kharga.  The main lithic activity was blade production.  Carbon 14 dates put this layer at around 29,950+/-900 (GrN-16782) and >- 30,000 (Lv-2084).

The most recent level, Level MP1, has a low density of artefacts which are distributed in a wide vertical scatter, and is clearly very different layers from MP2-5.  A few very characteristic tools are represented, principally two Emireh points, and burins on blades.  A few Levallois products are present and cores were all for intensively worked and used for blade production. The layer may be transitional between Middle Palaeolithic and Upper Palaeolithic industries. Van Peer et al (1996, p.153) suggest that the presence of the Emireh points, which have never before been found in any other African contexts, suggest south-western Asian contacts, and point to similarities with Boker Tachtit in the Negev, where the industry appears to be transitional between Middle and Upper Palaeolithic techniques. 

Van Peer et al (1996, p.153) conclude that “The older Middle Palaeolithic levels (MP5 through MP2) are clearly related to Middle Palaeolithic industries of the Nile Valley and the Western Desert (p.155), whereas Level 1 exhibits Near Eastern traits.

The presence of such a rich site, which was clearly re-occupied on different occasions over long periods of time, offers the hope that comparable sites will eventually be located in the Red Sea Hills.

A most Middle Palaeolithic drying of the area appears to have driven people from the Eastern Desert.  See the Environment section for more details. 

Upper Palaeolithic 35,000-21/20,000bp

The Upper Palaeolithic is a period of time during which many  writers see an occupation hiatus.  In Egypt as a whole very few sites date to this period, and this is probably due to environmental and climatic conditions. Extensive desiccation appears to have taken place in the eastern Sahara between c.60,000bp and 11,000bp, coinciding with the cold conditions of the last glaciation, and is marked by an extremely dry period identified at Dakhleh Oasis between 20,000 and 14,000bp. (Wiseman 2001, p.16). There are few Upper Palaeolithic sites in Egypt and Nubia, and most of these are restricted to the Nile valley.

Upper Palaeolithic tools have been found at Laqeita Oasis (Herbert and Wright 1988) and along the Red Sea littoral (Montenat 1986; Gawarecki 1986).  Although the El Gouna is well represented during the Acheulean and Middle Palaeolithic phases, there is no evidence for an Upper Palaeolithic occupation. 

The Sodmein Cave site mentioned above in the context of the Middle Palaeolithic occupation, also has two Upper Palaeolithic levels.  Level 1 is poorly documented, may have been disturbed, and may be a redeposited component of the second Upper Palaeolithic level which lies beneath it.  A few blades and cores have been found from Level 1.  Cores were discoidal and used for the production of flakes.  Level 2 was one of the richest occupation levels in the cave, and includes an important accumulation of lithics around a hearth.  It has been dated by AMS to 25,200+/-500 (UtC-3313).  The lithics consist of numerous blades, with unidirectional dorsal scar patterns present.  Cores feature one striking platform and were used for blade production.  Other tools were very rare, and there is no sign of the earlier Levallois technology.  The cave was abandoned until the Neolithic.

At Laqeita a knapping site was found, with retouched tools similar to types found in levels I and II of the Kom Ombo Sebilien

In the Nile Valley, the sites of Makhadma Sites 2 and 4, c.10km northeast of the mouth of Wadi Qena consisted of fishing camps dating to around 12,000bp coincinding with the Wild Nile stage of high Nile floods.  A the site a single shell of Engina mendicaria from Red Sea was found.

Late Palaeolithic 21/20,000 – 13,000bp

The Late Palaeolithic is well represented only in Upper Egypt where sites are dated to between 21,000 and 12,000 BP (Vermeersch 2002, p.35). Because of the lack of any detailed information about the Upper Palaeolithic, the origins of the Late Palaeolithic are uncertain. The oldest phases of the Late Palaeolithic, the Fakhurian and the Kubbaniyan are both represented at Wadi Kubbaniya near Aswan in southern Egypt and are apparently connected with the creation of a temporary lake in a wadi, which was maintained by the water table.  The sites were used by small hunter-gatherers who occupied the sites on an occasional basis each year for many years. The floral remains suggest that the sites were occupied on a seasonal basis. For more about the Late Palaeolithic of Egypt see the Late Palaeolithic page on 

There are no known sites from the Western Desert or Middle Egypt (Vermeersch, 2002, p.35). Likewise, although the Eastern Desert was occupied during Acheulean, Middle Palaeolithic and Upper Palaeolithic phases, there is no evidence for a Late Palaeolithic occupation, in spite of numerous remains representing several different industries in the Nile Valley.  Again, this may be an artefact of the shortage of surveys, rather than a real representation of the situation in the Eastern Desert at this time.

Wild Nile Phase c.13,000-12,000BP

At the end of the Late Palaeolithic, the Nile Valley experienced what is referred to as the “Wild Nile” phase – a period when the Nile saw huge fluctations in its levels between fast high floods and very low floods. This period, which lasted for around 1000 years only saw occupation in the Upper Egyptian Nile Valley, with no signs of settlement in the Eastern Desert, Western Desert or Middle Egypt.  Most sites were located on the desert fringes where the highest levels of Wild Nile floods reached (Vermeersch 2002, p.35), and indicate a riverine adaptation with seasonal fishing, plant collection and processing by grinding, and some hunting of large mammals.

Egyptian Occupation Hiatus c.13/12,000 – 8,500BP

Following the Wild Nile phase there appears to have been a period when Egypt was probably not abandoned in its entirety, but populations did move away in significant numbers (Vermeersch 2002).  There are no known sites from Upper Egypt, Middle Egypt, or the Western or Eastern Deserts.  Vermeersch suggests that “as the river level fluctuated wildly from year to year, the stability of the resource zones paralleling the channel deteriorated” (2002, p.35), with high flood levels causing the land to be unusable for long periods and low flood levels failing to provide the pools and marshes that were necessary for seasonal survival.  Vermeersch suggests that a possible competition for resources would have lead to situations which may be represented by the cemetery at Jebel Sahaba, where those buried met their ends in a very violent manner.  It is likely that any small site surviving have been buried by Nile alluvium. 

At around 10,500 the Western Desert saw some return of populations, with playa sites and hilly areas being occupied, but there are no known Nile Valley or Eastern Desert sites and only one possible date for this period from the Faiyum Qarunian (Vermeersch 2002) in Lower Egypt.

Epipalaeolithic .c8500-7500 BP

The origins of the Epipalaeolithic are somewhat uncertain, and appear to be somewhat discontinuous with previous periods.  It is possible that some Epipalaeolithic groups moved into Egypt during a period of post-glacial warming during the early Holocene, at the same time as significant changes in the Near East were accompanied by changes of lifestyle and economy, with increased sedentism. These innovations in settlement patterns and economic activities during the Near Eastern Natufian and the Pre-Pottery Neolithic B from around 9600-8000 BP were changing the face of the Levantine economy.  Playa deposits appear to have been laid down from around 12,000bp, suggesting that humid conditions began at around this time. 10th millennium bp sites are known from a number of areas in Egypt and the eastern Sahara, including the southern Western Desert, Siwa, Farafra, the Gilf Kebir, the Libyan Acacus and the Air Massif in Niger (McDonald 2001, p.28).  In the Western Desert, the desert areas were reoccupied from around 9500bp at Napta Playa (a hiatus of some 2500 years).  The  earliest occupation in other parts of Egypt are represented the Epipalaeolithic industries along the Nile valley at El Kab and in the Faiyum Depression (the Elkabian and Qarunian respectively), which begin at around 8000bp. For more information about the Epipalaeolithic of Egypt, see the Epipalaeolithic page on

The only known Epipalaeolithic site in the Eastern Desert is Tree Shelter.  Tree Shelter is located 1km away from Sodmein Cave.  It consists of a rock shelter near the mouth of a canyon-like wadi which has carved its way through layers of limestone.  The occupations are found in a small fan of sediments created by run-off waters (Van Peer et al 1996). Two occupation layers, one Epipalaeolithic (Lower) and one Neolithic (Upper) are present in the deposits of the shelter.  Unlike Sodmein Cave, the Tree Shelter deposits are largely non organic. 

The Lower, Epipalaeolithic, occupation probably represents a living floor.  It has numerous ochre features and produced some very rich archaeological material.  A preference was shown for the production of well made bladelets on local good quality red-brown flint, selected specifically from a range of other cherts, 3km from the site (Vermeersch et al 2002).  Tools are well represented, making up 400 modified items, most of which are very elongated backed blades and bladelets (28%) comprising a number of forms, including pointed straight backed, curved backed and partially backed and shouldered bladelets.  Other forms include shouldered points, ouchtata bladelets, some geometrics and microburins (18%) including elongated scalene triangles, notches and denticulates (22%) and endscrapers (18%).  Extensive use was made of the microburin technique, including the manufacture of finely denticulated bladelets.  Lithic microwear analysis indicates hunting and hide working activities.  Vermeersch et al (1996) are confident that “this assemblage can be attributed to the Elkabian, an Epipalaeolithic industry from the Nile Valley, that has been dated to around 8000BP” (p.417).  Elsewhere they describe the Elkabian industry and that of Tree shelter as “nearly identical” (Vermeersch et al 2002, p.125-9), with the only key difference being the presence at Tree Shelter of end-scrapers. This is also contemporary with Elkabian type sites in the Western Desert (e.g. Kharga and the Dyke area – Vermeersch 2002), which are accompanied by ceramics and cattle, and the Epipalaeolithic occupation (Qarunian) of the Faiyum Depression. Other artefacts included perforated ostrich shell beads, but no ceramics.   The Red Sea Elkabian does not include ceramics or cattle.  Mollusc remains include Nile Aspatharia and Red Sea species.

Vermeersch et al (2002) state that the archaeological data indicates several short occupations.  Dates span some 600 years, from 8120+/-45BP (KIK-656/UtC5389) to 7790+/-70BP (KIK655 / UtC-5388).

Vermeersch (2002) describes the occupation pattern of the Elkabian tool makers as that of “nomadic hunters, following eat-west routes with wintertime fishing and hunting in the Nile Valley and exploitation of the desert during the wet summer” (2002, p.36).

The extension of the Elkabian from the Western Desert, via the Nile Valley, into the Eastern Desert wadis indicated use by groups of more than one environment and region, taking advantage of benefits offered by each area.  The lack of cattle in the Red Sea assemblages may reflect different uses of landscape at different times of year for different herding purposes, or it may indicate different territorial regions within the Elkabian.

The Elkabian bears no similarities to the Qarunian Epipalaeolithic of the Faiyum Depression, to which it is clearly unrelated.


Where the Blue and White Niles meet a vast settlement area from the 7th millennium BC onwards was identified and excavated by Arkell in the 1940s. Arkell named it “The Early Khartoum.” The artefacts of the site are characterized by quartz flakes and brown incised pottery, grind stones and fragments of shell.  No earlier Pleistocene sites have been found which might have evolved into the Khartoum Mesolithic and at the moment its origins are a mystery.  The inhabitants of Khartoum Mesolithic settlements were describes as hunter-fisher-potters.  There are no signs of domesticated plants or animals, but the economy and technology had progressed from earlier Palaeolithic examples, so for convenience it has been designated Mesolithic.

There is no evidence in the Eastern Desert, at the time of writing, for a Khartoum-style Mesolithic phase.

Post-Epipalaeolithic Occupation Hiatus 7500-6900BP

Following the Epipalaeolithic there was, again, an occupation hiatus of several hundred years, with many areas of the Eastern Sahara being abandoned (Vermeersch 2002).  It is not clear how these movements out of the Sahara operated, or in which directions.  There is no evidence of occupation in Egypt at this time, apart from a single date from the Faiyum.


Later Prehistory

The Later Prehistory of Egypt defies the traditional conventional breakdown of the three-age system and its subdivisions.  The “Neolithic” is a complex phenomenon which, if loosely defined as the adoption of agriculture, appears in different forms at different times. 

Unlike the prehistoric Near East, where cereals were domesticated before animals, in Egypt domesticated animals were adopted before domesticated plants came into use.

Cattle may have been domesticated indigenously, although this is a matter of some debate, and cattle may have been introduced from the Near East.  In the Sahara, nomadic pastoralism based on cattle herding is associated with pottery.  These groups predate the use of domesticated cereals in Africa.  This is a reversal of the Near Eastern situation, where there were pre-pottery cereal growers.  The style of these Saharan ceramics is very distinctive and can be subdivided into a number of categories.  These groups who used the pottery are found over a vast area of northeast Africa and employed other types of artefact which differentiate groups on a regional basis.  Throughout the Saharan Neolithic, cattle were clearly revered, and for the first time cattle burials are found as well as human burials associated with cattle remains.

The earliest domesticates to unquestionably arrive in Egypt from outside Africa were goat and most probably sheep.  Sheep and goat were probably introduced from the Near East via the Eastern Desert into the southern Western Desert at Napta Playa and Dakhleh Oasis. 

Pigs appear at the same time as cereal domestication, in the Faiyum Depression and Merimde I at c.5500.  The precise nature of the occupation at these sites is not clear.  Although the components of the cereal and animal farming are known, there are no permanent settlement remains, field systems or evidence of burials. Connections at this time between these early agricultural communities and the Red Sea coast are suggested by the discovery of Red Sea shells in the Faiyum (Caton-Thompson and Gardner 1934) and Merimde (Hayes 1965), and slightly later at El Omari (Debono and Mortensen 1990).

There is a 900 year hiatus between the establishment of mixed agriculture in Lower Egypt and Upper Egypt.  It appears at c.4400 in the Badarian.  In Upper Egypt the Badarian form of mixed farming is accompanied by settlement and burial evidence, indicating a degree of sedentism.  Indications of mobility do exist, however, including the manufacture of portable items and the presence of bovids in burials.

As with Near Eastern prehistory, matters are very difficult to divide up with a simple terminology to deal with the adoption of domesticates in Egypt.  In the following Neolithic section I have divided the later prehistoric into an Earlier and Later Phase.  The Earlier phase includes the earliest Neolithic occupation of the Eastern Desert up until the Tasian, and corresponds to the earliest occupation of the Western Desert until the Faiyum Neolithic and the earliest phase at Merimde Beni Salama.  The Later phase starts with the Tasian and ends with the Badarian.


Earlier Phase 6900 – 5300BP

Two key sites make up most of the evidence for the earliest Neolithic in the Eastern Desert – Sodmein Cave and Tree Shelter Cave.

Sodmein Cave

Sodmein Cave, which has already been described in the context of the Middle Palaeolithic, contains Neolithic layers which are mostly organic in nature. The Neolithic reoccupation of the cave appears to have coincided with improved climatic conditions at around 7000bp. The faunal remains, in contrast to those of the Middle Palaeolithic, indicate climatic conditions that were only slightly better than today’s, with sparse rains and maybe an oasis or semi-permanent lake (Moeyersons et al 2002).  Similar conditions were found at Tree Shelter, only a few kilometres away from Sodmein.  For more details about the prehistoric environment see the Environment section.

Ten species of tree were identified, but were dominated by Acacia and Tamarix as they are today.  Present day wild fauna was accompanied by dorcas gazelle and Procavia capensis (rock dassie, which can no longer survive in the area due to the removal of acacia trees), several types of insect and some rodents. Domesticated species were confined to sheep/goat (ovicaprines), of which there were plenty of remains.  Enormous amounts of domesticated goat/sheep dung were found, and Vermeersch et al (1994) suggest that these layers indicate multiple visits by groups with large herds.  They believe that the ovicaprines were introduced from southwest Asia, and that they were present just after 7000bp (1994, p.39), predating their presence in Napta (6700bp) and Dakhleh (6500bp). 

Lithic artefacts and sherds are well represented throughout the Neolithic layers but only occasionally do they form clear occupation floors.  Ceramics were only found in the upper layers.  15 hearths were found at various depths.

1000s of lithic artefacts were discovered, made exclusively on local chert. Lithic tool production was largely opportunistic in character, with rudimentary techniques employed.  The cores show no sign of specialized preparation, and were mostly of single platform type, used to produce flakes.  Lithics were mainly flakes, with tools being rare.  Tools include retouched and notched flakes, poor quality side scrapers, some larger bifacial tools, rare blades and Microblades and a series of arrowheads, including some asymmetrical foliate forms, two stemmed points and some foliates, some of which are similar to those from the Bedouin Microlithic of Kharga Oasis, Dakhleh Oases Bashendi A and the Nile Valley Badarian, all of which dated to between 6500 and 5500bp. 

The ceramics in the upper levels are represented by a small number of sherds, made of a well fired red polished sandy fabric, varying in colour from dark red to brown, some with incised horizontal herringbone decoration.  The decoration was incised following polishing with a blunt polished tool.  The ceramic sherds are very closely comparable to Merimde’s level 1 “Fischgratverzierte Keramik” (dating to c.5900bp), which was atypical of the rest of the Merimde levels.  The main difference between the two decorative styles is that the herringbone pattern on the Sodmein sherds is rather more widely spaced. 

The 15 hearths found at different depths suggest repeated short visits by people and their herds (Vermeersch et al 1996).

Tree Shelter

Tree Shelter, mentioned in the context of the Epipalaeolithic, above, has also produced some Neolithic remains.  The Upper occupation layer is separated from the Lower Epipalaeolithic sediments by 0.8m of sterile deposits.  Considerable erosion is implied by the condition of the stratigraphy, and this may mean that only a small percentage of the archaeological remains have survived. 

Tree Shelter appears to have been reoccupied at the same time as Sodmein Cave, 1km distant from Tree Shelter, at around 7000BP with the onset of better climatic and environmental conditions.  It consists of numerous hearths, one of which was a small pit filled with charcoal and burnt cobbles. 

Animal remains are poorly preserved, but include Dorcas gazelle and domestic ovicaprines (2 goats). The earliest evidence for goat is associated with Feature 12 which dates to 6630+/-45BP (GrN-12560). Fish bones (some Scaridae) and Red Sea molluscs were present. 

Lithics, made on local chert cobbles derived from Eocene limestone, consist of a rough flake-based industry similar to the Neolithic industry at Sodmein Cave, but with few specific tools present.  The manufacturing technique was rough, leaving behind low quality debitage and cores with no consistent manufacturing technique employed. The best represented of the 80 implements are retouched flakes (27 pieces) denticulates (40 items or 50% of the assemblage), four bifacials, two arrowheads and 1 sideblow flake.  Vermeersch et al (2002, p.135) suggest that there is a similarity between knapping techniques of Tree Shelter and the Tarifian – and to a lesser extent to the Badarian site Maghar Dendra 2 (dated to 5110-5480BP).  Sideblow flakes are uncommon in Upper Egypt.  They are better known from Kharga Oasis, the Faiyum Depression, Dakhelh’s Bashendi Unit and the Western Desert Napta site E-75-8.

Carbon 14 dates from hearths indicate several occupation levels at this time. The earliest is 6770+/-60 (GrN-22562) and the most recent is 4930+/-30 (GrN-22651).

On the basis of the data from Sodmein and Tree Shelter, Vermeersch et al conclude (1996) that there was no dateable hiatus of occupation in the Sodmein area from 8120 – 6300bp, during which time 15 hearths were made at Tree Shelter and over 20 at Sodmein Cave.  Sodmein continued to be occupied until the onset of a dry phase at c.61200BP, but Tree Shelter was occupied until c. 5000BP.  They suggest that “During this period, people or ideas seem to move over a very large area from the Western Desert over the Nile Valley into the Red sea Mountains (p.135). 

Affinities between the Tree Shelter site and other Egyptian industries have been discussed in the Vermeersch et al paper (2002). They suggest that there is a similarity in knapping techniques to the Tarifian, and to a lesser extent the Badarian site of Maghar Dendra 2 which has been dated to between 5110 and 5480BP.  Side-blow flakes are rare in Upper Egypt – more are known from the Faiyum and Kharga where they are represented mainly as unstratified surface finds, in the Dakhleh Bashendi unit which dates to around 6000BP, and at the late Neolithic Napta Play site of E-75-8. Vermeersch et al (2002, p.135) conclude that “During this period, people or ideas seem to move over a very large area from the Western Desert over the Nile Valley into the Red Sea Mountains.  However, in an earlier paper Vermeersch et al (1996) suggest that the dates indicate an alternative, whereby ovicaprines were introduced from southwest Asia possibly via the Eastern Desert into the Nile Valley.

A possible “steinplatze” was found on a terrace above the wadi floor where the Wadi Bili meets a short wide wadi, at a site called ME03/10/24 inland of El Gouna.  Surface finds indicated that the terrace had potential, and six 40x40cm test trenches were excavated in an attempt to clarify the situation.  The site was dated to 5820+/-50bp (GrA-23727; Vermeersch et al 2005), which places it in the Middle Holocene.  Steinplatze are frequently recorded in the Sahara (Gabriel 1987) and are defined by a scatter of stones sometimes accompanied by big grooved stones, often interpreted as animal anchors. There are certainly rock art images which suggest that animals were anchored in some way via a tether attached to an ill defined floor level object.  Other frequently associated artefacts are grinding stones, although stone tools and ceramics are rare, and settlement structures or other features are also uncommon. They are often interpreted as temporary encampments for pastoral nomads. Site ME03/10/24 might be part of this tradition.  Its dates are also consistent with the Tasian site at Wadi Atulla, mentioned above, and a small shell mound. The archaeological data is only consistent with that of the Wadi Atulla site, and even then only to a limited degree.  If ME03/10/24 is accepted as a steinplatze site, then it is possible that Tasian cattle herders were exploiting conditions in the Red Sea Mountains between c.4800-4540 Cal. BC (Vermeersch et al 2005). 

Also in the El Gouna – Red Sea area a shell mound was discovered, dated by carbon 14 to 5800+/-40bp (GrA-23726).  However, the archaeological material from this site is completely different from that of any sites discovered so far in the area (Vermeersch et al 2005).

A somewhat controversial period designated as the Tarifian sits in this period, and has been mentioned above in relation to the Tree Shelter occupation.  The Tarifian is interesting because it sits at around 6300BP, therefore predating the Badarian period.  It predates domesticated animals or plants, but is associated with pottery production. The Tarifian has been discussed by Ginter and Kozlowski (1984) and has been summarized usefully by Diane Holmes (Holmes 1988, p.82) who has examined the lithic assemblages for affinities with the Badarian:  “The Tarifiarn appears to be a nonagricultural tradition . . . . Tarifian sites have produced pottery, but not wares which resemble any of the Badarian ceramic types. The lithic industry is a flake industry characterized by abundant endscrapers, notches, denticulates, perforators and retouched pieces. Even allowing for the fact that we are comparing modern excavated material with older collections, it is clear that the Tarifian industry is not the same as the Badarian though it is possible that it could be related to the Badarian in some way”.

In the roughly contemporary Faiyum A (Faiyum Neolithic), and the earliest phases of Merimde Beni Salama the presence of Red Sea gastropods suggests a connection with occupants of the Red Sea areas.

The Sodmein and Tree Shelter sites suggest two models for how the pastoral economy of Egypt developed.  The two models have been variously debated, both of which find some support in the data that has been available until recently.  The first model suggests that nomadic herding originated in the Near East and traveled south into the Western Desert.  The second suggestion is that nomadic herding originated in the Western Desert and traveled north into the Eastern Desert.  The El Gouna area, which has provided some evidence of Neolithic occupation, indicates that other sites are almost certainly waiting to be found, and it would certainly be interesting to explore the connection of any sites found with the Faiyum sites and Merimde Beni Salama.

Later Phase


A robbed tomb at Wadi Atulla to the north of the Wadi Hammamat (Friedman and Hobbs 2002) produced ceramics which can be compared with those identified as Tasian.  The Tasian was identified by Guy Brunton in 1934 based on the discovery of distinctive artefacts, particularly pottery decorated with incised geometric patterns, in the Badari region.  He suggested that it was the earliest Predynastic evidence for the use of ceramics, predating the Badarian and that the phase was characterized by round-based pottery made of calciform material decorated with incisions which were picked out with white pigment (also known from contemporary sites in Neolithic Sudan).  However, ever since Brunton’s work, there has been debate as to whether the Tasian actually existed or not. Although Brunton proposed that it predated the Badarian, some writers believe that it might have been a component part of the Badarian itself: “from the start the existence and chronological placement of the Tasian has been a subject of controversy” (Friedman and Hobbs 2002, p.178). Recent work in the Eastern Desert and elsewhere has suggested that the Tasian is more than a figment of the imagination: “The Tasian-related material in the Eastern Desert and now a site near Nabta Playa in the Western Desert . . . suggest that this culture may be the ‘missing link’ in the picture of interaction between desert dwellers and the Nile valley cultures, which led ultimately to the development of Egyptian civilization” (Friedman and Hobbs 2002 p.178)

Friedman and Hobbs (2002) working in the Eastern Desert at a site in Wadi Atulla (perpendicular to the Qift-Quseir road) have provided new insights into this period. The Eastern Desert has been conspicuous by its absence in accounts of prehistory and early Predynastic contexts.  The excavation of the Wadi Atulla tomb, was a salvage operation.  It had been plundered and damaged, and because of the lack of information about Eastern Desert sites its salvage was given a high priority. The tomb was built into soft sandstone. The excavators say that the tomb and its artefacts “give some indication of the original wealth and cultural milieu of its multiple occupants” (Friedman and Hobbs 2002, p.178).  It was visited briefly in 1983, when many of the plundered items which had no financial value were left to the side of the tomb entrance, and photographed by one of the Wadi Atulla excavators. These abandoned finds were almost totally depleted when it came to excavating, but the photographs provide information that is useful as supplemental data. Calibrated radiocarbon dates obtained from the site suggest that it dates to between 4940 and 4455 BC.  Although the site’s use as a tomb could well have been secondary (it may well have been a quarry site or a water gathering device) it was certainly used to deposit human remains. Although there were only a few undisturbed contexts, and bone remains were fragmentary and poorly preserved, the human remains of at least 14 individuals were found, varying in age from young (18 months) to middle aged.  It appears that these remains were deposited over a period of time, rather than as one mass burial at one time.

The ceramic forms from Wadi Atulla, with their incised decorations and clay compositions. appear to be made of local rather than Nilotic clays and are most similar to Guy Brunton’s Tasian examples. Further analysis of ceramics suggests that they were also manufacture locally.

Similar material to that of Wadi Atulla is found in the Western Desert at Gebal Ramlah near Nabtah Playa and may indicate connection between the Eastern and Western Deserts, and related pottery has found in the Nile Valley at a cave burial in Wadi el-Hol on the Farshut Road within the Qena Curve (Darnell, D. 2002), which may also indicate evidence for interaction between different distinct areas of Egypt.

Friedman and Hobbs believe that there are enough parallels between the Tasian identified by Guy Brunton and the Wadi Atulla remains to make the Tasian a valid phase:  “A number of shared traits certainly indicated a relationship between the desert and the valley and strongly suggests that the Tasian is distinct culturally, if not temporally, from the Badarian, as Brunton first maintained” (Friedman and Hobbs 2002, p.184). The beakers are the characteristic feature that suggest a link between the Nile Tasian and the Eastern Desert examples, although there are many differences which are important and indicate that the desert and Nile groups, while possibly in contact, were significantly different.

To see some images of Tasian ceramics, go off-site to the Digital Petrie address following: 


The Badarian is named after the site where it was first identified by Guy Brunton and Gertrude Caton-Thompson at el-Badari (near Sohag) and is represented by a number of small sites near the villages of Qau el-Kebir, Hammamiya, Mostagedda, Deir Tasa and Matmar as well as further a field at Mahgar Dendera, Armant, Elkab and Hieraknpolis and even in the Eastern Desert at Wadi Hammamat, with regional differences showing in different assemblages. At Hammamiya Caton-Thompson was able to identify a sequence of stratified levels that extended from the Badarian through to the Late Dynastic period.

The Badarian phase describes the earliest agriculture in Upper (southern) Egypt. It represents considerable change:  “With this culture we unexpectedly plunge straight into a symbolic universe of incredible richness, reflecting an increasingly structured and complex society, and this process was to accelerate enormously throughout the fourth millennium BC, eventually significantly contributing to the emergence of ‘Egyptian civilization’” (Midant-Reynes 1992/2000 p.152).  The Badarian definitely occurs before the Naqada culture, and may have been in existence by 5000 BC but “it can only be definitely confirmed to have spanned the period around 4400-4000 BC” (Vermeersch and Hendrickx 2000, p.40). It may have survived longer in the el Badari area than in other areas where the Badarian is visible, where it was probably contemporary with Naqada I.

The origins of the Badarian are something of a mystery.  Although the Levant is probably a likely origin for the appearance of agriculture there is nothing in the toolkit to suggest that there is any more of a connection. Holmes (1988) has examined Tarifian lithics to determine if there was a connection between the assemblages of the Badarian and Tarifian, but has concluded that there was probably not a close relationship, if one actually existed at all.  Stone tools are similar to those from the Late Neolithic in the Western Desert, and these seem strong enough to imply close links. The rippled pottery suggests connections with either the Late Sahara Neolithic or the Merimda or Khartoum Neolithic sites.  It is most likely that the Badarian was influenced by a number of different areas and concepts.

The evidence for the occupation of the Egyptian Eastern Desert during the Badarian is mixed. 

At Laqeita Oasis (Wadi Hammamat) pottery was dated by Debono (1951) to the Badarian and Naqada I periods.

In 1923, an expedition to the Red Sea coast by G.W. Murray (Murray and Derry 1923) located a recently plundered grave c.5 miles inland and due west of the Ras Samadai headland.  It was located on the terrace of a small tributary of the Wadi Samadai.  It was surrounded by a ring of large stones and had a roughly circular outline with a 3 foot depth.  Professor G.A. Reisner confirmed a Late Predynastic or Early Dynastic date for the burial, but more recently Resch (1964) has suggested that it is more consistent with the Badarian. The grave produced a skull and lower jaw, some decomposed leather,  a rectangular slate palette stained with green malachite, some fragments of a large grey coarse earthenware pot, a smooth quartz pebble, fragments of a tortoise-shell armlet, pierced shells, a bone awl, a wooden implement (tentatively described as a spoon) and a triangular fragment of quartz crystal with a worked point.  Derry points to the Nile-Red Sea connections suggested by frequent discoveries of Red Sea shells found at Naqada, Ballas and Lower Nubia.  The skull was considerably damaged but, in the estimation of Derry, was consistent with other Predynastic forms that he had examined.

Badarian burials frequently contain Red Sea shells (Brunton and Caton-Thompson 1928). Badarian graves occasionally contain large quantities of glazed steatite beads (e.g. tombs 5364, 5701 and 5733 – Brunton and Caton-Thompson 1928).  The steatite has not been sourced, but it is quite likely to have come from the Eastern Desert where it can be found c.160km from El Badari. Badarian groups also used agate and olivine, both of which can be found in the Eastern Desert and the Red Sea areas. Stone palettes were also found in graves for the first time.  They were made out of metagraywacke, which was sourced exclusively from the Wadi Hammamat in the Eastern Desert.  This is the stone that was used for palettes throughout the Predynastic period, and was known in later times as bekhen or bekheny stone:  “Its uniform fine-grained texture, and especially its green colour account for greywacke’s appeal over the millennia” (Harrell 2002, p.239).

The earliest examples of metal use come from the Badarian, where copper items are known, but these are very isolated examples, dating to after 4400BC (the earliest use of copper is in Anatolia at 6500BC with the next known date for metallurgy in Iran at 5500BC).    Tutundzic (1989) has suggested that the earlier appearance of copper in Upper Egypt at Badari than in Lower Egypt, where its earliest appearance is at the late Neolithic site of El Omari and the beginning of Naqada II is evidence of a route to Sinai via the Eastern Desert:  “It is hard to accept that transport of native copper by the Lower Egyptians, or via their territory was not accompanied by their own use of copper, in particular if such a transit was practiced over a long period of time” (p.257).  He sees the Upper Egyptian development of a direct connection with the Sinai peninsula either overland via the north-eastern part of Egypt or “more probably  Via the Red Sea route, or both ways” (1989, p. 259) as the most probable solution to the dilemma.  He also suggests that this linkage caused “the differences in the development between the Neolithic of Lower Egypt and of the contemporary Badarian culture in Upper and Middle Egypt” (1989, p.259).

The jury is still out on whether or not the Badarian communities were incoming groups or indigenous: 

Arguments in favour of incoming groups:

  • Caton-Thompson (1928) argued that the use of rough flint nodules, instead of fine chert available in nearby Eocene cliffs, indicated that the Badarians were unfamiliar with the surrounding landscape and its resources.
  • Arkell (1975) believed that the use of ibex in motifs, combined with the lack of harpoons in the Badarian assemblage, indicated that Badarian groups originated from the Red Sea Hills to the south of Qena.

Arguments in favour of indigenous groups:

  • Holmes (1989) examined the Badarian lithics as part of a much bigger study of Upper Egyptian lithics and concluded that the raw materials used were perfect for the tools being produced, and do not indicate a lack of knowledge of other materials.

Majer believes that some items reached the Badarian Nile settlements as a result of either collecting trips or exchange mechanisms: “Everything points to a similarity of culture between desert and valley with no temporal primacy for one or the other.  It is not a case of desert dwellers descending en masse into the valley from some Eastern Desert homeland (Kaiser 1956, 1985), but rather a long term interrelation between desert and valley in which movement between the two zones was relatively easy” (Majer 1992, p.228). 

Where Majer (1992) sees different groups interacting between Nile and desert/hills, Wilkinson sees a rather different pattern, where Badarian groups used both Nile and desert/hills in a subsistence economy based on both cultivation and seasonal pasturing of domesticated animals, combined with hunting.  He does not exclude the idea of desert-occupying groups, but does suggest a more complex set of processes than that allowed for by any single model (Wilkinson 2003).

In the southwest to southeast orientated Wadi Deir Bolos (off which the Coptic Monastery of St Paul is located, and where Middle Palaeolithic finds were discovered, described above), Late Neolithic finds are located on the main wadi deposits as well as on gravel at higher levels.  Dittman cites a number of examples.  Unfortunately, although he labels the layers, he doesn’t name the sites.  For the purposes of this paper, I will call it WDB-1.  Dittman says that it “consists of relicts in an abri which roof has fallen, and conservated parts of stratified layers” (1993, p.150).  There are five stratified layers of site WDB-1, which produced carbon 14 dates that have been calibrated to provide a picture of four cealry destinguishable occupation periods (1993, p.151):

  • WR-4
    • Earliest occupation period c.3630 – 3360 BC
    • western side of site
  • WL-1, WL-2, WM-1, WM-2 on the western side of the abri and OL-2 on the eastern side
    • Second period of occupation 3050-2690 BC
    • Numerous fragments of animal bones
      • Nubian wild donkey
      • ibex
      • Sheep/goat
  • OL-2 and OL-1 on the eastern side
    • Third period of occupation c.2660 – 2390 BC
    • Stratigraphic connections with WR-3 and WR-2,5 on the western side
  • WR-2 and WR-1
    • Fourth phase 2150 – 1920 BC
    • Last documentable phase of occupation



The evidence for the early Prehistory of the Eastern Desert is poor for the Lower Palaeolithic, but Sodmein Cave, to the north of Wadi Hammamat, has provided a colourful portrait of life during a humid Middle Palaeolithic, with a number of occupation phases coinciding with conditions that supported a number of plant and animal species that were adapted to these  aquatic and humid conditions.  Occupation patterns are sporadic during the rest of the Palaeolithic.  The Epipalaeolithic Elkabian appears to have ranged across the Eastern Desert, the Nile Valley and the Western Desert, showing that groups were adapted to a number of environments which they were able to exploit with considerable efficiency, altering aspects of their tool kits and herds to comply with the conditions that they were using.  There is no evidence for a Khartoum-type Mesolithic in the Eastern Desert.

The Eastern Desert has produced the earliest known evidence of “Neolithicization” in Egypt, with the appearance at Tree Shelter and Sodmein Cave of domesticated goat and sheep, clearly kept in small herds which could be moved according to the prevailing conditions. 

The Tasian period, which is still quite poorly understood, may have been a similar adaptation to that of the Elkabian, in the sense that it ranges across both Eastern and Western Deserts, but with the addition of pottery and a specific burial tradition.

The Badarian represents a considerable change in land usage bordering the Eastern Desert, with groups occupying the floodplain for cultivation and animal husbandry, but also some indications that the Desert was still used, both for obtaining greywacke and perhaps for animal pasturing by these mixed farmers.  Small settlements at the very edge of the desert may be the remains of temporary occupations by nomadic groups who used the desert on a permanent basis for herding and hunting.  It is possible that both types of group existed, employing different subsistence strategies in different situations, but interacting with each other on an occasional basis.


Go to top of page
Go to Predynastic page


Copyright Andie Byrnes 2007