Water resources and their management in the Eastern Desert of Egypt from Antiquity to the present day
Contribution of the accounts of modern travelers and early scholars (1769‑1951)
p. 451-492
Résumés
Since the last great period of aridification of the Sahara about 5,000 years ago, crossing or inhabiting the North African desert regions has meant being able to locate, maintain and protect watering points that are essential nodes in the exchange networks. These challenges are particularly important in the Eastern Desert of Egypt, which separates the Nile Valley from the Red Sea, because there are no perennial oases, nomadic populations have long escaped the control of successive states, and routes are constrained by topography. The region has nevertheless maintained a strategic importance, at least since the predynastic period and up to the present day, for its natural resources (particularly mineral and mining resources and, to a lesser extent, vegetation) and for its essential place in travel and trade networks (Red Sea ports and access to the African coast and the Indian Ocean, pilgrimages to Mecca). Since the voyage of James Bruce in 1769, many European and North American travelers and explorers have traversed the Eastern Desert of Egypt and produced accounts and maps, the critical analysis and cross-checking of which make it possible to better assess the ways in which water was managed from the 18th to the 20th centuries. They thus allow us to document very precisely the period before the introduction of motorized vehicles and hydraulic pumping and to establish comparison points with earlier periods and with current networks. When combined with archaeological and ancient textual data in a regressive approach, these elements open up new perspectives for the diachronic study of the Eastern Desert watering points.
Depuis la dernière grande période d’aridification du Sahara il y a environ 5 000 ans, traverser ou habiter les régions désertiques d’Afrique du Nord signifie pouvoir localiser, entretenir et protéger les points d’eau qui sont autant de nœuds essentiels dans les réseaux d’échange. Cette maîtrise est particulièrement importante dans le désert Oriental d’Égypte, entre la vallée du Nil et la mer Rouge, car la région ne comporte aucune oasis pérenne, les populations nomades y ont longtemps échappé au contrôle des États successifs et les routes qui la traversent sont contraintes par la topographie. La région a néanmoins occupé une importance stratégique, au moins depuis la période prédynastique et jusqu’à nos jours, en raison de la présence de ressources naturelles (notamment minérales et minières et, dans une moindre mesure, végétales) et pour sa place essentielle dans les réseaux de voyage et de commerce (les ports de la mer Rouge donnent accès à la côte africaine et à l’océan Indien, et sont essentiels pour les pèlerinages à La Mecque). Depuis le voyage de James Bruce en 1769, de nombreux voyageurs et explorateurs européens et nord-américains ont traversé le désert Oriental et ont produit des récits et des cartes dont l’analyse critique et le recoupement permettent de mieux apprécier les modes de gestion de l’eau du xviiie au xxe siècle. Ils permettent de documenter très précisément la période précédant l’introduction dans la région des véhicules motorisés et du pompage hydraulique et d’établir des points de comparaison entre les périodes antiques et les réseaux actuels. Associés à l’analyse de données archéologiques et textuelles antiques dans une approche régressive, ces éléments ouvrent de nouvelles perspectives pour l’étude diachronique des points d’eau du désert Oriental.
Entrées d’index
Mots-clés : désert Oriental d'Égypte, période gréco-romaine, voyageur moderne, gestion de l’eau, puits, réseau d'approvisionnement, qalt, zone humide, source
Keywords : Eastern Desert of Egypt, Greco-Roman period, modern traveler, water management, well, supply network, qalt, wetland, spring
Texte intégral
Introduction
1Since the onset of hyperarid conditions in the north-eastern Sahara around 2500 BC,1 knowledge and control of watering points and routes have been of major importance to inhabitants and those who exploit desert resources. During this whole time span, the Eastern Desert of Egypt was exploited through mining (gold, copper, etc.), quarrying (granite, porphyry), husbandry and charcoal making. It was also crossed by caravans and expeditions from/to the Red Sea, linking the Nile and the societies of Egypt and the Mediterranean to the rest of the region (Arabian Peninsula, sub-Saharan Africa, India), making the Eastern Desert of Egypt a critical region in the ancient networks from the 4th millennium BC onward, and the gate to the Red Sea from Egypt and the Mediterranean Sea (fig. 1).2 As a result, the archaeological sites and remains of occupation are numerous and form a dense network of hundreds of sites, from simple rock shelters, to mining or quarrying settlements and fortresses that have hosted thousands of people for decades and even centuries.3
Fig. 1 – Map of the Eastern Desert showing the main ancient and modern roads and settlements (L. Manière/Desert Networks).

2Since 2017, the ERC funded project Desert Networks4 started to work on Eastern Desert connectivity from the New Kingdom to the end of the Roman period (ca 1500 BC‑ca 300 AD). The reconstitution of the physical networks involves, in addition to the study of the network nodes (i.e. the archaeological sites), working on the links between the nodes: it is necessary to reconstruct the conditions of journeys in the desert during Antiquity, by assessing the location and chronology of the archaeological sites known in the region, including the watering places, which are some of the main and critical nodes, the paths and the transportation means,5 alongside the geography and geology.
3This contribution aims to document the different water resources that ancient travelers, on their route to the Red Sea, had at their disposal. Two main challenges arise: there has been considerable modification to the environment and water management after the introduction of motor pumps in the region from the 20th century; the ancient written sources on water availability are scarce, sometimes difficult to interpret or indirect, and archaeological data are almost non-existent.
4To overcome these difficulties, this article will be based on the examination of more than 60 accounts dated mainly from the 18th‑20th centuries and produced by western travelers and early scholars, who traveled in the region in almost the same conditions as in ancient times, as we will show in the first part of this contribution. We will also examine the modern watering places mentioned in modern topographic maps.
5We intend to assess the types of watering places mentioned in modern documentation, their operating conditions, the related practices and their role in modern desert networks. By doing so, we will produce data on water uses and on previous states of the resources and structures, thus providing additional information to archaeological6 and historical7 studies.
6First, we will try to verify if the typology which emerges from the processing of travelers’ accounts can be extrapolated to Antiquity,8 and is relevant in the field and to the terminology used in written sources. Secondly, we will see if similarities between Antiquity and modern times are valid for all periods. Finally, we will examine the connectivity between the Eastern Desert hydraulic networks as they appear to have functioned in Antiquity and modern times with the other adjacent hydraulic networks, in the Nile Valley and on the Red Sea shore.9
7This discussion will take into consideration the archaeological and written sources at our disposal, and will often focus on the Ptolemaic period, on which our project is centered.10 We will rely on recent data drawn from the field by the French archeological mission of the Eastern Desert (MAFDO).11 We will also rely on the thousands of ostraca uncovered by the mission in a dozen Roman and three Ptolemaic forts and on the archaeological data their excavations yielded.12
Background
Geographical context
8The area considered in this paper is located between 27°20’ N and 23°15’ N, from the Red Sea to the Nile Valley (see fig. 1). From west to east, this desert consists of two distinct topographic units, an extensive sedimentary plateau in the west, and the mountainous Red Sea Hills in the east. In the northern part, the topographic surface of the plateau corresponds mainly to a geological formation consisting of Eocene limestones; in the south it is mainly composed of Upper Cretaceous sandstones (Nubian sandstone). The altitudes rise from west to east, and the wadis are increasingly incised towards the east of the plateau. This incision is further deepened in the Red Sea Hills where the divide between the Mediterranean and Red Sea watersheds is located. This mountain range is made of igneous rocks and metamorphic rocks that are intensely fractured and faulted. None of the watercourses of the Eastern Desert are perennial13 and most of them flow only during rare episodes of intense rainfall.
9Since the end of the African humid period, between 4000 and 3500 BC, the climate of the Eastern Desert of Egypt has become more and more arid14 until becoming hyperarid around 2500 BC.15 The mean annual rainfall value in this part of the Sahara is now less than 5 mm.16 At the same time, because of high temperatures and strong winds the mean evaporation rates now reach between 4,000 and 6,000 mm per year.17 Vegetation is thus scarce and located only in specific places (parts of the wadis where the water table is close to the surface, north-facing slopes where the evaporation rates are lower, coastal areas). Thanks to previous less arid periods and to the hydrogeological configuration of this area, there are some groundwater resources, some of which are finite – they are thus decreasing over time. Some others, often smaller and situated near the surface, are renewed by rain events and/or underground circulation. Two main types of underground water compartments are found in the area: wadi alluvium aquifers and rock aquifers.
Corpus and method
10The typology of water resources was produced from the systematic analysis of accounts of travelers and modern scholars (18th‑20th centuries), and the study of current geological or topographical maps and satellite images.18 Information from travelers’ accounts was compared with topographic and hydrogeological parameters and with the location of all wells referenced on the 1/50,000 (Egyptian General Survey Authority, 1989‑1990) and 1/250,000 (Army Map Service Corps of Engineers U.S. Army, 1953‑1960) topographic maps.19
11The Desert Networks database includes a total of 252 located water points (fig. 2) and one that is not yet located.20 Most of them are found in the Red Sea Hills. The Arabic word bi’r is the most commonly used to designate them, but it can be used to designate both a well (hydraulic structure) and a larger area where a well can easily be dug because the water table is close to the surface. These may even be areas located in natural basins formed in the geological substrate of wadis, which are filled with gravel and therefore need to be drained in order to access water or even natural basins collecting surface water (qilāt).21 All the water points referenced were observed on satellite imagery, which made it possible to determine that 6 of them correspond with qilāt, 6 of them are springs and 68 correspond to wells. The remaining 172 water points could not be assigned to a type because no structure was visible. The number of natural basins is underestimated here, because most of them are hard to see on satellite imagery.22
Fig. 2 – Map of the watering points of the Eastern Desert recorded in the Desert Networks database (L. Manière/Desert Networks).

12The corpus of modern and contemporary sources brings together data from 24 trips made between 1769 and 1951 (tab. 1).
Tab. 1 – List of the journeys and travelers’ accounts used to draw the typology of the Eastern Desert water resources.
Studied journeys | ||
Date | Participants | Bibliography |
1769 | Bruce | Bruce 1790 |
1792 | Browne | Browne 1799 |
1799 | De Rozière and Denon | Denon 1802; De Rozière 1801‑1802, 1813 |
1799 | Bachelu | Dubois‑Aymé 1802; De Rozière 1801‑1802 |
1800 | Bert | Couyat 1909, 1910 |
1818 | Belzoni | Belzoni 1820; d’Athanasi 1836 |
1823 | Wilkinson | Wilkinson 1832 |
1841 | L’Hôte | L’Hôte 1841 |
1850 | Du Camp and Flaubert | Du Camp 1860; Flaubert 1910 |
1873 | Colston | Colston 1886 |
1886 | Floyer | Floyer 1887 |
1890 | Nicour | Raimondi 1923 |
1899 | Mac Alister | Mac Alister 1900 |
1907 | Weigall | Weigall 1913 |
1922 | Bisson de la Roque | Bisson de la Roque 1922 |
1816‑1818 | Cailliaud | Cailliaud 1821 |
1831‑1832 | Linant de Bellefonds | Linant de Bellefonds 1868 |
1864‑1875 | Klunzinger | Klunzinger 1878 |
1888‑1889 | Golénischeff | Golénischeff 1890 |
1897‑1898 | Barron and Hume | Barron, Hume 1902 |
1902‑1908 | Ball | Ball 1912 |
1907‑1955 | Murray | Murray 1925, 1955 |
1930‑1935? | Scaife | Scaife 1935 |
1947‑1951 | Tregenza | Tregenza 1955 |
Inheritances and continuity of practices in the Eastern Desert
13The study of the corpus highlights some continuity in the exploitation of certain water points, linked with the routes used by travelers from the 18th to 20th centuries which often corresponded to ancient routes.23 The maintenance of major poles at both ends of some routes helps to explain this continuum. For example, the ancient Coptos‑Myos Hormos road, the road through the Wadi Hammamat, still connects the Red Sea port of Quseir on the Red Sea near Quseir al‑Qadim,24 which corresponds to the Ptolemaic and Roman port of Myos Hormos,25 to Qena, which took over from Coptos,26 the starting point of the road during the Roman period. In the 19th century it was known as the “Russafa road” and was one of the main roads linking the Nile Valley to the Red Sea.27 Thus, even travelers who did not intend to visit or study archaeological remains traveled on it, such as for example Browne in 1792,28 Bachelu who went down the road during the French Campaign in Egypt in 1799,29 and Du Camp and Flaubert who followed most of it during their journey in 1850.30 Similarly, the ancient route from Coptos to Berenike was followed over a large part of its course by Colston when he was ordered to join Purdy at Berenike from Cairo in 1873; the soldier traveled up the Nile to Qena, from where he went deep into the desert, and described many ancient stations along the way.31 The expeditions of Cailliaud32 and Belzoni33 – in which d’Athanasi also participated as interpreter – followed the road from Edfu to Berenike between the Nile Valley and the fort of Abu Rahal. They used the ancient stations and roads, well known to their Ababdeh34 guides,35 as landmarks.36
14Along the same lines, the use of camels from the first half of the 1st millennium BC,37 and their widespread use in caravan transport from the beginning of the Ptolemaic period38 partly explains the continuity of water management and well location. From the 18th century until the 1950s, travelers used camels for their journeys (with the notable exception of Bruce in 1769, who traveled mainly on horseback, and part of his caravan shared this means of locomotion despite the presence of camels).39 The comparison with the Greco-Roman period is thus easier than with the New Kingdom, when donkeys were the pack animals employed.40
15Another factor of stability is environmental:41 in a stable climatic context on the scale of the last 5 millennia, the areas favorable to the constitution of surface or underground water resources have changed little because they depend mainly on criteria that were relatively unaffected by human activities before the 20th century (hydrogeological configuration, topography, etc.). The means of exploiting these resources have also changed little between Antiquity and the introduction of motorized pumps and deep drilling means. Shadufs42 have been used since the 2nd millennium BC43 and the oldest remains of sāqiyat44 in Egypt date back to the Late Ptolemaic/Early Roman period.45 This explains why it is possible to observe the maintenance of certain water points over long periods of time, the wells of certain ancient stations were still exploitable in the 18th and 19th centuries46 and some still contain water today.47 It was a common practice among desert dwellers to search for the location of ancient wells during the 20th century, and reopen them.48 Ancient wells still in use, or reused after being dug or deepened, were numerous in the 19th century Eastern Desert and new wells were in fact often dug close to old wells. For example, it has been said that a well dug by Seti I in the Wadi al‑Sidd was reopened in 1905 to supply the workers of nearby gold mines,49 which were also exploited in Antiquity.50 It was unlikely that this was in exactly the same location, but it was a well exploiting the same water resource. Similarly, in al‑Kanais, a well was successfully reopened in 1905 (it is still visible; fig. 3),51 not far from an inscription engraved with the name of the pharaoh Seti I which claims that a well was dug in this locality to supply travelers.
Fig. 3 – Picture of the modern well at al‑Kanais (G. Pollin 2018/Ifao/MAFDO).

Water points and exploitation of their resources
16The corpus of 18th‑20th century travel narratives that was analyzed made it possible to assess the different types of water points in the Eastern Desert, and their exploitation. A typology was constructed, solely based on the accounts of the travelers. It includes six main types: wadi; wetland; qalt; spring; surface water catchment area; well. Built cisterns are not included in this paper because we did not find any mention of these in use between the 18th and 20th centuries. The contribution of this to the interpretation of archaeological or ancient textual data will be developed in the discussion and conclusion.
Wadi
17The water of wadis (fig. 4) is rarely mentioned in the accounts of travelers. And for good reason, it is rare that they contain water, as they depend exclusively on surface water, making this difficult to observe. Nevertheless, wadi flows were observed by Klunzinger near Quseir52 and by Scaife at the outlet of the Wadi Qena in March 1934.53 In the same year he observed flows in the Wadi al‑Qirayyah, several days after a storm.54
Fig. 4 – Picture of Wadi Jirf near Xeron Pelagos after a rain event (M. Reddé 2010/MAFDO).
18These flows do not necessarily produce water reserves that can be directly exploited on the surface, except in rare overflow basins or when the course of the wadi is interrupted by a natural or man-made obstacle (fig. 5), but they do contribute to recharging alluvial aquifers, and to the development of vegetation useful for pastoralists. Tregenza cites Wadi Umm Araka as an example: “there were plenty of zille bush for the camels. He [Soliman, a Maaza travelling with Tregenza] had found water by digging down a meter in the sand”.55 Bert found water under the same conditions in Wadi Hawashiya in 1800, again the vegetation here was highly developed. They can also be used by water intakes (see below).56 Moreover, it is also possible to cultivate with flood water.57
Fig. 5 – Satellite image of a dry temporary lake in the Eastern Desert, 10 km north of the Paneion in Wadi Menih al‑Hayr; the black line indicates the maximal extension of the lake (M. Crépy, WGS84, UTM36N; map base: BingMaps).

Wetland
19Wetlands have the dual interest of constituting a consumable water resource (at least for animals) and pasture that is easily renewed. One is mentioned at Bir Ambagi/Ayn al‑Ghazal. Brought about by a spring, it is also the starting point of a small stream. The area is well documented because it is located at an important crossing point, where at least four roads connecting the Nile Valley to the Red Sea meet between Bir Beida/Bir al‑Ingliz and the port of Quseir. The descriptions are therefore particularly rich. De Rozière who visited the place during the French Campaign in Egypt in 1799 specifies for example:
C’est à deux lieues et demie de Qoçeyr qu’on rencontre la dernière source; elle est entourée d’une végétation fort abondante, comparée à la nudité absolue des environs. Ce lieu, connu sous le nom de Lambâgeh, est un des plus remarquables de la vallée, et le seul qui offre un site agréable. […] Au milieu coule un ruisseau d’une eau très limpide, mais qui, dans la saison des pluies, se change quelquefois en un torrent considérable. […] L’eau de Lambâgeh sert à abreuver les chameaux des caravanes ; mais les hommes se gardent bien d’en boire, car elle passe pour très malsaine: elle m’a paru seulement douceâtre et un peu pesante à l’estomac; qualités qu’elle doit au terrain gypseux sur lequel elle coule.58
20This area is also mentioned and described by Du Camp,59 and Flaubert60 who visited the site together in 1850. Klunzinger, who lived for several years in Quseir, also described this wetland:
The springs that here and there occur have a very bitter taste, and sometimes give out a smell like that of sulphuretted hydrogen. In places the soil appears loose, crusty, yellowish, moist, as it were, spongy, and impregnated with a saline fluid. A bitter, perennial rivulet, the Ambagi, makes a vain attempt to trickle farther down into the valley, and gives a verdant existence to a grove of rushes, but after a few days’ rain becomes a raging, devastating stream.61
21This wetland still existed later and was described by Barron and Hume62 and by Weigall.63 Today, 16 ha of wetlands and vegetation are still visible on satellite images (fig. 6) at Ayn al‑Ghazal.64 The sustainability of this waterhole, and its proximity to Quseir al‑Qadim, the ancient Myos Hormos (8 km by foot, 6.5 km as the crow flies), makes it an excellent site for Pliny the Elder’s Ainos or Tadnos Fons.65 No ancient remains are visible near the site although this is not contradictory to Pliny’s testimony, since a fons does not have to be built.
Fig. 6 – Satellite image of the wetlands at Ayn al‑Ghazal (M. Crépy, WGS84, UTM36N; map base: BingMaps).

22Another spring-fed wetland, of much smaller extent, existed in Abu Shaar al‑Qibli,66 which hosted a Roman fort and an outside well probably of the same date.67 The site was long identified as Tadnos Fons mentioned above, due to the wrong identification of the nearby site, Abu Shaar, as Myos Hormos.68 No ancient name can be associated with Abu Shaar al‑Qibli, and the analysis of satellite images shows that it no longer exists and has now been replaced by cultivated land.
Qalt
23A qalt is a natural basin (sometimes slightly modified by human intervention) with impermeable or low-permeability rocky bottoms (fig. 7a‑7b) that fill with water during rainfall (concentration of local run-off only for endorheic depressions, storage of flood water for depressions in the bed of wadis) or that sometimes benefit from recharging from temporary or permanent springs.69 Except in the latter case, it generally does not contain water permanently, due to evaporation. However, when a qalt is sufficiently sheltered, it constitutes a water resource for several months. References to such features in the travelers’ accounts are so numerous that it is not possible, in the present state of work, to produce an exhaustive list. However, the corpus studied allows us to describe their functioning and use through examples. The earliest mentions in our documents correspond to Bruce’s travels from Qena to Quseir in 1769 and Bert from Asyut to Gebel Ghuwayrib, then Wadi Tarfeh and back to Asyut in 1800. Bruce mentions a night’s rest in a place called Mesag el‑Terfowey70 – masaq meaning “small rock basin usable by people or animal”71 – and describes their water supply as follows:
This water does not appear to be from springs, it lies in cavities and grottos in the rock, of which there are twelve in number, whether hollowed by nature or art, or partly by both, is more than I can solve. Great and abundant rains fall here in February. The clouds, breaking on the tops of these mountains, in their way to Abyssinia, fill these cisterns with large supplies, which the impending rocks secure from evaporation.72
Fig. 7a – Rocky threshold forming the downstream end of a qalt at Umm Rus (M. Crépy 2019/MAFDO).
Fig. 7b – Satellite image of one of the qilāt in Umm Disah, at a time when it had water; the dashed line indicates the maximal qalt area (M. Crépy, WGS84, UTM36N; map base: BingMaps).

24Bert mentions qilāt on two occasions. When he reached Gebel Gharib, near Umm Dhaya and Qalt Gharib, his guides found water in sanded qilāt (it was necessary to remove the sand to recover water).73 Later, in Wadi Tarfeh, famous for its water-rich qilāt, they were unable to refuel because all the qilāt were exhausted.74 In the same logic, Barron and Hume, at the end of the 19th century, thought of refueling in the Gebel Gharib in a place called Bir Gharib, corresponding to the unnamed qalt used by Bert’s expedition near Umm Dhaya; they did not manage to draw much water from it, in spite of having removed the sand and mud that was there.75 Despite the risk of finding temporarily depleted qilāt, these resources were considered relatively reliable and rather abundant in the 19th and 20th centuries. Wilkinson, for example, describes three of them that are particularly well supplied, at Gebel Hawashiya,76 at Gebel Gharib,77 and near Qattar, in the immediate vicinity of Naqqat.78
25Further south, Mac Alister, when traveling from Daraw to Gebel Sikait pointed out numerous qilāt (filled only after rainfall) near Sikait79 and also found an important qalt in Umm Selim.80 Finally, on the basis of Nicour’s notes, Raimondi, in his projects for the development of the Eastern Desert, in particular the setting up of a railway line between Kom Ombo and Berenike, evaluated the supply of water from the qilāt of the region for the supply of the construction site and then the railway line.81 One of them, which we were unfortunately unable to locate, still contained 100 m3 when Nicour visited in 1890, after 10 months without rain.82
26This reliability is based on the long-term sustainability of some of the qilāt. Tregenza thus mentions the sustainability of surface water in the Gebel Umm Anab83 and the importance of the waters of the Gebel Qattar84 for the camels left to graze freely, who can access water without human assistance.85 The latter region is well documented, especially from the end of the 19th century as it contains several well supplied qilāt (including that of Naqqat mentioned above) and many qilāt in an area called Qalt Umm Disah, first mentioned by Floyer who spotted it during his journey in 1886 and referred to it as Medisa glen:
It is a stiff climb down of four hours from the ridge to the Medisa glen, where is always a plentiful supply of water, both in a natural reservoirs with steep sides and full of green watercresses, and from holes scraped in the gravelly bed above. […] If, however, you follow the windings of the Medisa ravine, you will pass other large basins, one especially large one overgrown with calamus, or Arab pen-reed, which had only dried up in the summer of 1886.86
27Barron and Hume also mention that there was water there87 and add that the surrounding hills contain the best water resources in the region.88 Bisson de la Roque also saw water there in 1922.89 There was still water there at the time of Sidebotham’s passage in August 1997,90 thus testifying to the great durability of this water resource still mentioned on maps under the name of Bi’r Umm Dīsī. This does not mean that water will always be there, but that the probability of finding water there, even after long periods without rain, is high.
Spring
28The travelers also mention a large number of springs, i.e. natural water outlets, some of which may have been improved by humans. Although most of the springs do not seem to be suitable for exploitation today, it appears that they were sometimes useful in the past. It is unclear whether this is due to environmental change (water scarcity), changes in practices, or both. Some of them were indeed sufficient to maintain wetlands and to reinforce qalt recharge (see above). Others were sufficient to maintain micro-oases, for example at Bir Abu Safa, where Linant de Bellefonds observed one in 1832.91 This spring is located in the immediate vicinity of a façade carved into the rock dating back to Ptolemy III Euergetes,92 more precisely 228‑227 BC.93
29Similarly, the Amusne spring (see above) seems to have been of great importance, as it allowed entire groups to get water and stay on site for several days.94 On the contrary, some springs located on the road from Laqeita to Quseir used by the French Campaign in Egypt in 1799 did not have a sufficient flow to supply everyone,95 but in total this was an expedition of 1,000 or 1,100 people, each with a camel.96 It appears, however, that most of the springs produced low or even intermittent flows, as their recharge was generally local.97
30The study of the hydrogeological context of Laqeita, as well as the data provided by early scholars98 indicate that the wells of the site probably exploited ancient artesian springs benefiting from the waters of the Nubian Sandstone Aquifer System. In any case, the wells there benefited from artesian recharge, which explains the profusion of water available, which made this site strategic in Antiquity and in the 18th‑20th centuries,99 and still allows its agricultural development today. Since this aquifer has progressively drawn down over the last 7,000 years,100 water was closer to the surface in Antiquity than it is today, and it is even likely that it naturally outcropped or gushed.
Surface water catchment area
31More rarely, travelers mentioned improved or built basins and/or water intakes.
32The only type of water intakes that were observed in the Eastern Desert by travelers and early scholars, and only during the 20th century, were in the form of dams and constructed bypasses designed to divert flood water to closed depressions. Tregenza explains this as follows: “By building a dam of wadi-boulders and gravel he had trapped a local flow of rainwater and with it had flooded a wide bay of sand in the outer hills. He had thus been able to sow water-melon seeds and millet over an area of several acres, and his people were now harvesting the crop”.101 Current satellite imagery shows that these methods are still in use (fig. 8a), although dams made from wadi alluvium now appear to be largely replaced by trenches or levees made with backhoe loaders (fig. 8b). However, there are few observations, and it seems that these methods of development have been, at least for recent periods, very localized.
Fig. 8a – Satellite image of temporary agricultural fields in the Umm Yessar area, using wadi diversion by a wall; the green dashed line shows the maximal extension of the agricultural area, the black one indicates the location of the three main remains of a wall which is likely to be ancient as it has stood there long enough to modify the morphology of the alluvial fan, i.e. probably for centuries or millennia; wadi flow comes from the east (M. Crépy, WGS84, UTM36N; map base: BingMaps).

Fig. 8b – Satellite image of temporary agricultural fields in the Umm Yessar area; the green dashed line shows the maximal extension of the agricultural area; the black dashed lines indicate three backhoe trenches and levees used to divert water from the wadi; the arrow indicates the main wadi flow direction (M. Crépy, WGS84, UTM36N; map base: BingMaps).

33Apart from these few and recent examples, the other mentions of reservoirs built in the wadi correspond to misinterpretations of remains. The first example is Colston describing two circular stone tanks in the wadi at Daghbag (anc. Compasi), that were 9.50 m in diameter.102 It was in fact much more likely that he observed ancient quartz mills, used by gold miners.103
34The second example is excavations in the bed of the wadis, surrounded by thick and high piles of debris and whose functioning has been compared to that of the ancient Sudanese aḥfār (sg. ḥafr, or more commonly hafir, see below). A dozen of these features are known of in the Eastern Desert, all dated to Antiquity. The only traveler to mention them is Bisson de La Roque, who describes the so-called aḥfār (he named them “réservoir d’eau” and “prise d’eau”) at al‑Saqqia and Ghozza, a Roman fort and a Ptolemaic mining site.104 We will return to this type in the discussion since in fact most of them, if not all, are not aḥfār, but protected wells. Klunzinger, who lived for several years in the desert wrote about this: “Cisterns, in the sense of rain-water being directed into a pit and kept there for years, do not exist in this desert”.105
Well
35The last type of water resource is also the most frequently mentioned in the corpus and the most used by travelers. The typology presented here is based on the location settings, and not on the kind of resources exploited nor on the technologies involved. Two main types of wells are envisaged: wells protected from flooding (by the enclosure of a fort – fig. 9a and 9b –, by their topographical position106 or by the disposition of excavated and/or dredged material – fig. 10a, 10b, 11a and 11b) and unprotected wells.
Fig. 9a – Satellite image of the Deir al‑Atrash fort (M. Crépy, WGS84, UTM36N; map base: BingMaps).

Fig. 9b – Picture of the Deir al‑Atrash fort (M. Crépy 2020/MAFDO).

Fig. 10a – Satellite image of the hydraulic structure at Ghozza; the dashed line shows the general orientations of the rubbles heap protecting the well; wadi flows come from the east (M. Crépy, WGS84, UTM36N; map base: BingMaps).

Fig. 10b – Picture of the hydraulic structure at Ghozza (M. Crépy 2020/MAFDO).
Fig. 11a – Satellite image of the hydraulic structure at al‑Saqqia; the dashed line indicates the general orientations of the rubble heap protecting the well; wadi flows come from the north and north-east (M. Crépy, WGS84, UTM36N; map base: BingMaps).

Fig. 11b – Picture of the hydraulic structure at al‑Saqqia (B. Redon 2019/MAFDO).

36Among the protected wells, the most documented example is the series of ancient wells protected by the enclosure of Roman forts located along the ancient road leading from Coptos to Myos Hormos in Antiquity (from Qena to Quseir at the time of the accounts of travelers and early scholars), and reused/reopened during the 19th and 20th centuries.
37Most of the wells of the road were not functional at the beginning of the 19th century according to reports by Bachelu used by De Rozière, from the French Expedition to Egypt:
[…] ce sont des especes de caravenserails ou de mansions fortifiées, toutes construites à-peu-près sur le même plan. […] Ces quatre corps de bâtiments enferment entre eux un espace carré de dix-huit ou vingt toises de côté, libre de toute construction, mais dont le centre est occupé par un puits circulaire d’un diamètre considérable, autour duquel descend en hélice une rampe fort large, destinée autrefois à conduire jusqu’au niveau de l’eau. Actuellement ces puits sont en partie comblés.107
38During the 19th century, under the reign of Mohamed Ali and sometimes thanks to British engineers, several of these wells were reopened. Wilkinson indicates water in Bir Hammamat on a map published in 1832.108 In 1838, L’Hôte describes the wells of Laqeita and Bir Hammamat109 and mentions some other wells on the road. He describes Bir Hammamat as follows: “Ce puits est remarquable par sa profondeur et sa construction, analogue à celle du puits de Joseph, dans la citadelle du Caire. On y descend par un escalier qui tourne en spirale et qu’éclairent des jours pris sur le puits”.110
39In his book published in 1923, Raimondi mentions the following wells linked with ancient forts: Laqeita,111 Bir Hammamat which he wrote had been built under Mohamed Ali’s reign and was depleted when he came there,112 and Bir Sayala, equipped with stairs.113
40In the same area, wells are mentioned at Bir al‑Sidd114 and their situation seems to be favorable to their long-term maintenance, because they were located aside the wadi, near the head of a small watershed.
41The wells of Laqeita, located on the same road are a little bit different, built (with stones or bricks), but with no internal stairs but ramps. They are still in use nowadays, and are described by many travelers. The most complete description is from De Rozière:
[…] on y trouve trois puits, dont l’eau, fort abondante, a un goût plus désagréable encore que celle de Birembarh; elle n’est pas sensiblement salée et n’incommode pas. Ces puits, tous très larges, sont maçonnés intérieurement, et paroissent encore en bon état; un ou deux ont une rampe douce par laquelle les chameaux descendent jusqu’au niveau de l’eau, où se trouvent des espèces de réservoirs destinés à les abreuver; on est ainsi dispensé d’élever l’eau jusqu’à l’orifice des puits.115
42The last type of protected well is the one that Bisson de La Roque wrongly identified as a ḥafr. These hydraulic structures are wells, probably equipped with ramps and thus accessible to animals, and enclosed by rubble intended to protect them from wadi flooding, and from alluvium infilling.
43In addition to the protected wells, some wells depicted by the travelers were not protected, or not sufficiently protected. For example, the well (now infilled) of the Ptolemaic fort of Bir Samut (fig. 12), the enclosure of which was broken due to wadi flooding, has been described by several travelers in different ways: Belzoni mentioned the well, but did not specify if the well was infilled;116 Golénischeff thought the well was an infilled cistern, only visible by the abundance of grass in the middle of its courtyard;117 Murray described an open well, 20 m deep.118
Fig. 12 – A: Belzoni’s drawing of the plan of Bir Samut (1820); B: Golénischeff’s drawing of the plan of Bir Samut (1890); C: satellite image of Bir Samut (BingMaps).

44More generally, Ball indicates that “the wells of South-Eastern Egypt are mostly shallow excavations in the alluvia of the wadi floors” and that “the shafts dug by the Arabs are generally wide and crooked, in order to permit of a man descending to fill a water-skin. Usually a well consists of three or four such shafts sunk in proximity”.119 But the most important information he reports is on the evolution of these wells and the practices related to their operation:
After every considerable rainfall the wells become filled up with stony downwash, and have to be dug out afresh. There is no protective wall to prevent infilling; and, contrary to what might at first be thought, it is not laziness which conditions this circumstance. To the Arab, wells are the last resource. After rain, all the galts [rock basins] are full of good water. The Arab knows that the supplies in these galts will evaporate, while those in the wells, covered in by alluvium, are safe from loss by this cause. He therefore draws his supplies from galts as long as he can, and only when these are empty does he open the wells. The main wells never fail except after unusually prolonged drought, and then the condition of the Arab is sore indeed.120
45This quotation exemplifies the strategy put into place by the Bedouins in their water management, playing on the different types of resources and their characteristics, a strategy that included natural filling of the wells during flooding to protect the water in them, and their reopening when other resources were exhausted. Not only does this explanation help us understand why Bir Samut has been seen in such diverse states, but it also explains why few ancient wells have come down to us apart from those in the major forts and stations which still have their enclosure (at least partly) intact, and why so many of the water points mentioned on the maps do not appear on satellite images.
46On the contrary, some wells seem to be permanently open and are mentioned by many, if not all travelers, for example Laqeita, or Bir Abbad. This is not surprising since these two localities, situated close to the Nile, share the same hydrogeological configuration: they draw their water from the Nubian Sandstone Aquifer System, which is rich in water and conducive to artesian conditions leading to water rising towards the topographic surface. Thanks to their generally abundant supply over a long time, they had a particular status for travelers as they served as meeting points and staying places for caravans during their preparations and gatherings.121
Discussion
Is the typology based on travelers’ accounts relevant to Antiquity?
The rare occurrences of ancient use of water from the wadis, qilāt, springs and wetlands of the Eastern Desert of Egypt
47For the modern travelers, the wadis did not constitute a resource that was really exploitable as surface water, but rather potential recharges for the water tables of alluvial deposits and exploitable pastures, in particular by local populations. From the Old Kingdom to the end of the Roman period, under similar climatic conditions, the situation was probably not different. Nevertheless, the presence of gabions and walls protecting the Ptolemaic village of Ghozza against floods through the armoring of the wadi terrace on which it was implanted shows that the wadis’ flows were taken into account during Antiquity.122 Beyond the consideration of risk and the remarkable adaptation of techniques to the hazards of flooding that they emphasize (the village has remained almost intact), these elements demonstrate that the wadis could flow at that time, thus creating local surface water resources, recharging wadi alluvium aquifers and facilitating the establishment of crops, even in the middle of the desert. The potential contribution of wadi floods and rain events also appears in the “ostracon of the miracle” found at Bir Samut, in which a storm and a flood that contributed to the replenishment of a water point located nearby (Takha) are mentioned, and where the need to come and get this water to fill the cisterns is pointed out. This Demotic ostracon highlights practices of maximum exploitation of unanticipated water resources.123
48Qilāt can provide large quantities of water, without effort, and thus water large caravans or herds. But despite their reliability, finding water is never a certainty and they are often located far away from the main wadis. They therefore appear, in the modern accounts, more favorable to exploitation by local and mobile populations. In Antiquity, no major site was built near a qalt, except maybe the Roman fort of Badia, in the Porphyrites district, which is a little off the Qena‑Abu Shaar road.124 However, some qilāt were exploited, as shown by scattered pottery or inscriptions found in their vicinity.125
49Four ostraca from the fort of Umm Balad, in the northern part of the desert, near the Porphyrites, mention “the well of the watercress bed” (τὸ Καρδαμήτου ὕδρευμα) in the vicinity.126 Floyer’s descriptions of the qalt at Umm Disah includes a mention of watercress,127 confirming that this aquatic plant can grow in the heart of the Eastern Desert. There used to be small ruined buildings near the qalt128 and the location (23 km from Umm Balad, via the site of Qattar) and the description of Qalt Umm Disah makes it possible to identify this area with the “well of the watercress bed” mentioned in the ostraca. This would confirm the long-term reliability of this qalt and the role of Gebel Umm Disah’s surface water from Antiquity to our days as an important water node.
50Springs constitute a third water resource that is generally not exploitable. Most of these were locally recharged springs and were probably already of little use in Antiquity. Only those connected to aquifers of larger volumes could have been exploited over a longer period of time, as in the case of Amusne, that we have unfortunately been unable to locate for the moment, or Ayn al‑Ghazal. This last example is important since it is likely to be the Tadnos Fons of Pliny (see above): it is located 7 km south-west of Quseir al‑Qadim/Myos Hormos, and results from a local hydrogeological configuration which make this spring perennial on an historical scale. The source was particularly striking to travelers, as it gives birth to extensive wetlands, almost unique in the Eastern Desert. The water quality was judged as bad by most of the travelers, but they did not drink the water from the spring directly, but rather from the swamps.129
51Part of the underground resources of the Eastern Desert were inherited from wetter periods, so it is possible that better supplied springs existed during Antiquity. The remains of springs observed not far from ancient sites should therefore be evaluated by a systematic geomorphological assessment in order to qualify the potential resource. For example, during a MAFDO survey, a now dried-up spring and a qalt were located in the immediate vicinity of a small site (Umm Rus) made up of around 20 huts, dated from the New Kingdom.130 The location of this cluster of huts, far from the associated gold mines and in a narrow and steep wadi, was probably influenced by the water resources.
52Another good example of the changes that have occurred since Antiquity is illustrated by the sources about Laqeita. As mentioned above, this site had an important role as a gathering and supply point for caravans and expeditions before they entered into the desert. In Roman times too, Laqeita, called Phoinikon (a name which indicates that palm trees were grown in the small oasis-like site), played that role, thanks to its location at the beginning and crossing of two roads, leading from Coptos to Myos Hormos, and Coptos to Berenike.131 The site was not chosen by chance but certainly because of the presence of water in abundance. Moreover, this abundance was undoubtedly more important in Antiquity than in modern times, because of the drop of the water table level in the Nubian Sandstone Aquifer System. The artesian nature of the springs – a feature which is certain, but that has escaped most travelers and scholars because of the diminution in artesian power since Antiquity132 – was another asset for the site.
53Another consequence of the drop in the water table level is that the wetlands associated with certain springs were probably more numerous and extensive in Antiquity. However, outside periods of scarcity, the consumption of their water was probably reserved, as in more recent times, for animals. No such resources are mentioned though, to our knowledge, in Eastern Desert ancient sources or known by archaeology.
54It is possible that wetlands, but also wadi floods and diversions, qilāt, springs such as the one mentioned above, may have contributed to supplying water to gardens and crops (fig. 13) documented by ostraca in the vicinity of certain Roman forts.133 Several gardens are known of thanks to ostraca, the most famous being located near the forts of Umm Fawakhir/Persou, Laqeita/Phoinikon and Daghbag/Compasi. These small plots have never been seen in the field (the sites have been badly damaged), but the ostraca show that they were cultivated by soldiers and civilians settled on the site, with a lot of species such as vegetables and herbs.134 In these sites, the analysis of satellite images indicates potential cultivation areas, relying solely on underground water at Laqeita, or possibly also exploiting the diversion of wadis flows and the concentration of run-off near the two others.
Fig. 13 – A modern desert garden near the Pharaonic temple and the Ptolemaic fort of al‑Kanais (G. Pollin 2018/Ifao/MAFDO).

Reassessing the use of the so-called aḥfār in the Eastern Desert of Egypt
55The fifth type of water resource, water intakes, is rare in the modern sources. Before trying to find out if such facilities were used in Antiquity, some clarification is required in relation to a frequent misunderstanding of some remains, identified in the bibliography as ḥafr135 (tab. 2), in reference to hydraulic structures attested in Sudan during Antiquity.136
Tab. 2 – Eastern Desert structures identified as aḥfār in the previous bibliography.
Site | Desert Networks website | Identified as possible ḥafr or water intakes |
Abu Midrik | DN_SIT0017 | Krzywinski, Seland 2007; Sidebotham, Gates‑Foster 2019 |
Abu Shaar al‑Qibli | DN_SIT0028 | Sidebotham, Gates‑Foster 2019 |
Al‑Khashir | DN_SIT0032, DN_SIT0251 | Sidebotham, Gates‑Foster 2019 |
Al‑Saqqia | DN_SIT0097 | Bisson de la Roque 1922 |
Bayzah | DN_SIT0049 | Sidebotham, Gates‑Foster 2019 |
Bir Samut | DN_SIT0064, DN_SIT0395 | Krzywinski, Seland 2007; Sidebotham, Gates‑Foster 2019 |
Ghozza | DN_SIT0115 | Bisson de la Roque 1922; Sidebotham, Gates‑Foster 2019 |
Hitan Rayan | DN_SIT0123 | Aldsworth, Barnard 1996; Sidebotham, Gates‑Foster 2019 |
Rawd al‑Buram | DN_SIT0172 | Sidebotham, Gates‑Foster 2019 |
Rawd al‑Liqah | DN_SIT0174 | Sidebotham, Gates‑Foster 2019 |
Rawd Umm al‑Faraj | DN_SIT0176 | Krzywinski, Seland 2007; Sidebotham, Gates‑Foster 2019 |
Wadi Menih al‑Hayr | DN_SIT0117 | Sidebotham, Gates‑Foster 2019 |
56A ḥafr aims to harvest water, by trapping part of the run-off from hillslopes, or from a wadi flood after rainfall, in a dug basin surrounded by rubble heaps from the digging. It is therefore reliant on the mean annual rainfall and the distribution of rainfall during the year, on the location and the orientation of its opening(s). For maximal efficiency, rainfall needs to be quite heavy, and a rainy season is best, as it will ensure the basin is filled with water, and the collected water can be used during the dry season. A lot of them can thus be found in Sudan, especially in the region where the mean annual rainfall is between 50 and 300 mm,137 whereas the mean annual rainfall is lower than 5 mm in most of the Eastern Desert of Egypt. Considering this, the use of hillslope run-off would be very unproductive in Egypt, unless one builds a complete and extensive system for water harvesting and diversion, which does not appear to have been done anywhere in the Eastern Desert. This type of hydraulic structure is also extremely unsuitable in Egypt because of high evaporation rates. The observation of several attested Sudanese archaeological and/or present-day aḥfār on satellite imagery shows that the characteristic of any usable ḥafr should be the following:
- location in a wide wadi, with an extensive upstream watershed;
- situation in low areas of wadis, not on the uppermost terraces;
- orientation of the opening(s) facing upstream or perpendicular to wadi flows, never facing downstream;
- sediment infilling with wadi alluvium alternated with clay layers (stagnation of water).
57This being said, we shall move on to the analysis of the dozen structures identified in the Eastern Desert by some scholars as aḥfār.
58It is impossible to discuss the nature of the structures at Abu Shar al‑Qibli, Bayzah, Hitan Rayan, Rawd al‑Buram, Rawd al‑Liqah and Rawd Umm al‑Faraj: what is visible on satellite imagery is unclear and we did not survey these sites. On the contrary, we have visited Abu Midrik, Bir Samut, al‑Saqqia and Ghozza. At Abu Midrik, a Ptolemaic fort on the Edfu-Berenike road, both the situation and the orientation of the openings do not fit with an identification as a ḥafr. At Bir Samut, a Ptolemaic fort on the same road, the sediments are definitely linked to successive flooding inside the well of the fort since the destruction of the north-western corner and not to the voluntary digging of a ḥafr. In al‑Saqqia (see fig. 11a‑11b), a Roman fort on the Porphyrites road, the large depression south of the fort has never contained water from the wadi:138 it is situated on an upper terrace, has no opening and its sediment infilling is of aeolian origin. Accordingly, this structure is more likely to be a protected well whose rubble heaps prevent wadi floods from filling the well with gravels and pebbles.139
59At Ghozza, a Ptolemaic gold-mining settlement, the structure is located on an upper terrace, and has its openings downstream (see fig. 10a‑10b).140 Test pits showed that it was equipped with a ramp coming from the west. The sediment infilling of the ramp is not linked with wadi flow/flooding, and is mainly made up of aeolian sediments slightly reworked by local run-off and of coarse sediments that have collapsed or been displaced over short distances from the edges of the ramp. Its characteristics are thus not compatible with the ḥafr hypothesis, but with a protected well141 equipped with a ramp dug in the wadi alluvium.
60The last two sites will only be studied through satellite imagery. Al‑Khashir is located near the confluence of two wadis, but has been located under the protection of a high terrace between the structure and one of the wadis; thus, its water cannot reach the structure. Moreover, it has no opening, and even if we suppose that the thinner part of the rubble heap is more recent, the opening would have been oriented towards downstream parts of the wadi. The structure was made to prevent water from coming inside rather than for collecting it. The so-called ḥafr in Wadi Menih al‑Hayr is located on the floor of a natural temporary lake formed by a wadi whose flows are cut by a stable sand dune, and water is concentrated and stocked there naturally. The opening of the structure is located downstream. It remains unclear if it is really a kind of ḥafr, different to those from Sudan, which would have been a deepened part of a natural rainwater lake142 or if it is a protected well tapping into the water table generated by the lake.
61To conclude with water intakes, those described by Tregenza (cf. supra) which divert wadi flows to water gardens near Umm Yessar and Umm Disi, visible on satellite imagery, could have been used during older periods, as reported by Harrell143 and as suggested by the structures seen near Umm Yessar on the satellite images (see fig. 8a).
Wells: the main water resources in Antiquity and modern times
62Wells constitute the main water resources mentioned by the travelers. They are the only hydraulic fact mentioned in ancient written sources, besides cisterns (see below, on the terminology).
63Wells were obviously crucial to the establishment of the roads that allowed caravan crossings, and mining or quarry sites for the exploitation of desert resources. All route construction began with the digging of wells. The inscription of Seti I at al‑Kanais demonstrates this particularly well:
He said: “How painful is a way that has no water! What are travelers to do to relieve the parching of their throats? What quenches their thirst, the home land being far and the desert wide? […] I will make for them the means to sustain them, so that they may bless my name in the future […]”. Now after his majesty had spoken these words to his own heart, he went about on the desert seeking a place to make a watering station. […] Stone workers were ordered to dig a well in the mountains, in order that it might uplift the weary and refresh the heart of him who burns in the summer heat (transl. Lichtheim 1976, pp. 52‑57).
64More than 1,300 years later, under Vespasian, the dedicatory inscriptions of three Roman forts (and probably more) newly built along the Coptos‑Berenike road, at Siket, Didymoi and Aphrodites Orous, state that the prefect of Egypt, L. Iulius Ursus, ordered a well to be dug in the area, and then, for a fort and cisterns to be built:144 “In the 9th year of Imperator Caesar Augustus Vespasianus, L. Iulius Ursus, prefect of Egypt, returning from Berenike gave instructions for a well to be sought in this place. When it had been found, he ordered a fort and cistern to be constructed, under the direction of M. Trebonius Valens, prefect of the desert region of Berenike” (Siket inscription, transl. Bagnall 2005, in O.Ber. II 120).145
65As shown by these examples, wells were sometimes prior to the establishment of other facilities (road, forts, stations). Their location was chosen with care, and it is not uncommon for a site to be located on or near an older site where water was known to be available. Thus some ancient wells could be reused by a road, even though they were established to fulfill another function, e.g. the al‑Kanais well, intended to facilitate the exploitation of gold mines located to the east of the site during the New Kingdom, which became a well on the road from Edfu to Berenike and the Red Sea during the Ptolemaic period.146 The reoccupation of ancient sites is particularly common during that period like at Bir Samut, whose Egyptian name (if the name Proembesis is indeed to be attributed to Bir Samut)147 probably indicates a previous occupation under or near this fort that acted as a major logistical node on the same road.148
66The digging of wells was carried out under the responsibility of the army in Roman times.149 It was done by specialists in all periods,150 who could still encounter difficulties.151 Their maintenance was crucial and evoked many times by the sources.152 These regular dredging operations have left traces on the ground, notably at Xeron Pelagos where enormous layers of dredging material covered the oldest layers of the fort’s dump153 or at Abbad, where the center of the ancient Ptolemaic fort, notably the two cisterns, is occupied by layers of dredging resulting from the reopening of the well during the Late Roman period (fig. 14).154
Fig. 14 – Kite photograph of the fort of Abbad; the dredging material of the well occupies a wide area in the center of the fort (A. Rabot, G. Polin 2017/Ifao/MAFDO).

67The ostraca found in the desert show that some wells gave a lot of water155 or, on the contrary, were poor in water,156 and we even have a poem found at Xeron, and dated to the 3rd century AD, which lists the forts of the Coptos‑Berenike road, with details of the quality and taste of the water that their wells provided to travelers.157 This important aspect (water is sometimes unfit for human consumption, but not for animal consumption) is regularly mentioned in modern maps158 and travelers’ accounts.159
68Only a few ancient wells are known in the field, such as the two wells of the imperial porphyry quarry of the Porphyrites160 or that of Abbad (the only well located in a fort that has been properly excavated, although not completely). It is not the place here to draw up a typology, but it can be said that their management was very similar to that described by the travelers, with opening/reopening of wells when needed. This is true for wells protected by the enclosure of a fort. This is probably even truer for isolated wells or wells located outside enclosures, which were likely to be filled faster and which have totally escaped our knowledge (so far, the only ancient wells that have been studied are protected wells).
69One may notice that there is in fact a difference between Ptolemaic and Roman times. Indeed, during the Roman period, wells were dug at the beginning of (if not a little earlier than) the construction of forts, and because their occupation was apparently long-term and with no long period of abandonment, their maintenance was permanent (most of the forts of the Coptos‑Myos Hormos road were inhabited by soldiers and civilians for more than a century and the forts of the Coptos‑Berenike road for 150 years or so).161 Of course, the wells were likely to fill up, for many natural reasons, but this filling up was not fast since the well was constantly used and maintained.162
70On the other hand, in the Ptolemaic period (and certainly even more so in the Pharaonic period), the occupation of forts and roads was probably temporary, interspersed with relatively long episodes of abandonment.163 In this case, the filling up of wells was faster. One might wonder whether this explains the fact that the crowns of dredged material from the reopening of wells (as at Abbad, for example) are sharp and not removed from the center of the fort or spread out, to allow the location of the well to be found during future occupations.
71Before concluding on typology, the discussion on aḥfār (cf. supra) highlighted the existence of little described, or previously misunderstood, ancient structures: wells protected from floods by a heap of excavated material. At al‑Saqqia, al‑Khashir, and Ghozza (and perhaps at Wadi Menih al‑Hayr if it is indeed a protected well), in particular, their very large size and the work of digging and forming the rubble heap that they required, as well as the careful choice of their position, make them intriguing works. What were they used for? Why was so much effort taken to protect them? Without excavation and extensive fieldwork, it is impossible to say more about Wadi Menih al‑Hayr, al‑Saqqia and al‑Khashir. The MAFDO’s work in 2020 in Ghozza allows us to determine the following elements concerning the protected well located south of the Ptolemaic miners’ village:164
- the round excavation corresponding to the well is more than 20 m wide;
- the structure includes a steep ramp (more than 3 m wide at its westernmost part) oriented towards the center of the well;
- the ramp is fully enclosed by the protective rubble heap;
- Ptolemaic potsherds found on top of the rubble heap indicates that the structure is Ptolemaic or older.
72The presence of the ramp and the width of the well shaft, which implies a significant amount of digging work, suffice to explain why such care was taken in positioning the excavated material around the structure, to avoid rapid filling with coarse sediment during floods, which would require complex re-excavation, and could lead to massive collapses. The two pits dug in the structure in 2020 indicate that it worked, as most of the infill consists of fine sediment, which is easy to extract, and whose displacement by flow is not likely to damage the structure.
73There is still a question about the structure of Ghozza: why dig such an imposing shaft and build such a ramp? The comparison of the archaeological context with two lesser-preserved parallels at Semna and the Semna mine sites165 may give some answers. The three sites are located in the north of the Qena/Quseir axis, were occupied during the Ptolemaic period and were mining sites, for which water needs were significant, for the supply of water to workers and pack animals, as well as for gold processing.166 At a time when sāqiyat did not yet exist, it seems possible that these sites used ramped protected wells either to allow the installation of chains of shadufs or to allow animals (donkeys or camels) to go down to wells to drink and/or to bring water back to the site, thus accelerating the supply of water compared to all the other systems available at the time.
The terminology of water resources in the Eastern Desert
74The modern accounts show the variety of types of water resources that were at the disposal of the travelers. It is likely that most of them already existed in Antiquity. In the ancient sources dealing with water resources in the Ptolemaic and Roman Eastern desert though, the only word attested is ὕδρευμα (hydreuma), in addition to dexamene or lakkos167 i.e. “tank”, which are not discussed here.
75The term hydreuma is specific to the Greek language of Egypt.168 According to H. Cuvigny,169 it is used in the Eastern Desert ostraca in the sense of “well”, and never means a “tank” as it is often used in the Valley and western oases. This is shown by an example found in an ostracon (O.Claud. I 2), that mentions that Rayma’s hydreuma is being dug and the well-diggers have just reached the water table.170
76The word hydreuma in the ancient sources refers to the wells themselves. It has also been used to forge several toponyms of the Eastern Desert, mainly in the Ptolemaic or the Early Roman period. These include three forts with a well mentioned by Pliny the Elder (HN, 6, 102-103) at the beginning of the Roman period along the Coptos‑Berenike road: “Apollonos hydreuma”, “Trogodyticum hydreuma” and “Novum hydreuma”. It is meaningful that the other stops on Pliny’s list (which came to him from Eritrean trade merchants) are listed as “a hydreuma”, where the caravans could be supplied with water, or as a place “in monte” (in the desert), for simple stops, with no constructed or perennial or semi-perennial water points.171 The fort built in the Early Ptolemaic period at the site of al‑Kanais is referred to as “the hydreuma near the Paneion” (τὸ ὕδρευμα ἐπὶ τοῦ Πανείου) in an inscription engraved nearby (I.Paneion 12, l. 5‑6, 232 BC). Other hydreuma place names appear in the Greek ostraca of Bir Samut (dated to the 3rd century BC), among the main stations known so far in the Ptolemaic road system,172 including the site known as “Apollonos hydreuma” in Roman times, although with no explicit reference to the hydreuma in its name.173 The hydreuma-named places are thus a Ptolemaic phenomenon, while in Roman times, the forts of the two main roads are called praesidia (forts). This illustrates the importance of water availability, which is the raison d’être of the construction of the road stops in the Ptolemaic period, called by H. Cuvigny “the period of the hydreumata”, in opposition to “the time of the praesidia” of Roman times,174 during which security concerns were more critical.175 However, the word hydreuma does not disappear from our documentation in Roman times of course, and water and water resources are still evoked in the ostraca (see above, “Wells: the main water resources in Antiquity and modern times”).
77Despite the lack of variety of water management in the written sources, archaeological evidence shows that, in fact, wells are not the only way to collect water in the Eastern Desert. One then should ask if the term hydreuma could have had a broader sense and designate a water point that has been set up so that people and beasts can refuel there. In ostraca and inscriptions directly dealing with the Eastern Desert, other broader meanings have already been proposed in some cases, such as “water from the water table” or simply “water”.176 If the location of the Καρδαμήτου ὕδρευμα is indeed at or near Bir Umm Disah like we suggest, this would show that a water point set up near a qalt to facilitate the cultivation of watercress in the Eastern desert could also be called a hydreuma.
The water networks of the Eastern Desert of Egypt
78We have been focusing so far on the water points of the Eastern Desert. It is now time to look at the networks these nodes have formed.
79In terms of the organization of water networks, a shift appears in the modern accounts, starting at the end of the second quarter of the 19th century:
before that date, the trips (e.g. Belzoni’s journey or Cailliaud’s second expedition, which faced serious lacks of water)177 were made with little water, a systematic and sizable supply before traveling, and a constant fear of running out of water, which had to be found by reopening wells whose locations were only known thanks to local guides; they also relied on springs and qilāt, two less secure and more ephemeral water resources;
- after the construction and reopening of ancient wells in the years 1830‑1850 by the British army, British entrepreneurs and the Egyptian administration, the journeys were a little more signposted (e.g. the journey of Du Camp and Flaubert), and water management was easier thanks to the networks of wells, newly built with red bricks, stones and mortar; even if the risk of running out of water does not disappear, it is less prominent in the accounts.
80Similarities can be found here with Antiquity, the first kind of network being close to the Ptolemaic one, while the more recent network looks more similar to the Roman (post-Vespasianic) one.
81In Ptolemaic times, the network was composed of two kinds of nodes: nodes related to the roads leading to the Red Sea, and nodes related to the exploitation of gold and emerald in the area (see fig. 1). Even if exploitation was also at work in Roman times (of hard stone, porphyry and granite), this was limited to a very defined area, north of the region (Mons Claudianus and Porphyrites), and quite far from the two main roads leading to the Red Sea, from Coptos to Myos Hormos and Berenike. This was not the case during the Ptolemaic period, when gold and emerald mines were close to the roads and spread over the entire desert surface, from north to south.178 Thus the Ptolemaic network of sites, in terms of topology, was much more meshed than the somehow linear networks of the Roman period, at least if we exclude the northern part with the quarries.
82The other difference between the two networks is the fact that the occupation of the Roman sites, as mentioned above (“Wells: the main water resources in Antiquity and modern times”), was continuous over a long period of time, while the Ptolemaic occupation was more sporadic, and evolved geographically over time. Consequently, the occupation took a different form, with large Roman praesidia equipped with wells located between 20 and 30, maximum 40 km from each other to form a tight mesh, with everything built not all at once but over a relatively short period of time and following a more or less established plan, whereas the Ptolemaic period mesh was looser, based on a few forts, some small stations and rock shelters, a network that sometimes presented large gaps with no wells or other water sources available. They formed thus two kinds of networks: an intensive one (with few nodes, but which were strong and important) and an extensive one (more extended over the desert, but with less strong nodes).
83Water management along the roads was by consequence different. In Roman times, in particular after the great building program undertaken under Vespasian,179 the water supply does not appear frequently in the ostraca. H. Cuvigny goes so far as to say that the supply of water to the caravans in the direction of the Red Sea was not the main role of the forts, all the more that they were built when the harbors of Berenike and Myos Hormos were in decline.180 She proposes that their construction, and the strengthening of the network between the Nile and the Red Sea under Vespasian was mainly related to security matters; the fact that the forts were built at short intervals was convenient for the post office and the army supply caravan, which were made up of donkeys, and not to supply the trade caravans. Although demonstrated by much evidence, this hypothesis does not necessarily negate the fact that water was still crucial for the people living in the forts (and the harbors) to survive. Moreover, donkeys’ water needs are high. Securing a water supply was thus crucial for the maintenance of roads and their functioning, and the Roman administration had therefore taken care to provide inhabitants and travelers (whoever they were) with an abundant and secured supply of water at regular intervals. The equipment of routes with protected wells was done in several waves, first under Augustus181 and then under Vespasian (with later adjustments).182 The low frequency of occurrence of water-related issues in the ostraca could thus be linked to a judicious initial choice of well locations, which ensured their efficiency and abundance throughout the period of road utilization.183
84In Ptolemaic times, the security of the water supply was more uncertain. This was clearly stated first by Strabo (17, 1, 54):
Now in earlier times [i.e. the Ptolemaic period and maybe also the Pharaonic one] the camel-merchants travelled only by night, looking to the stars for guidance, and, like the mariners, also carried water with them when they traveled; but now [i.e. under the Roman rule] they have constructed wells, having dug down to a great depth [bathos], and, although rain-water is scarce, still they have made cisterns for it (transl. Jones 1932).
85This has been dramatically demonstrated by the discoveries of 17 ostraca in the Ptolemaic fort of Abbad, which is the first station on the main road of that time, between Edfu and Berenike.184 Here, the expedition of Lichas, who led two elephant hunting expeditions to the Horn of Africa, gathered to be supplied with food, mainly bread,185 and water. The ostraca mention more than 360 people were recruited for the expedition (soldiers, guides, donkey drivers, hunters) and the amount of water to which they were entitled. It recalls the situation of Laqeita, where modern travelers and also the Napoleonic army, gathered to constitute caravans and troops, and refuel with water, before leaving for the desert (see above). Bir Abbad, located 2 km to the west of the Ptolemaic fort of Abbad, is also mentioned in modern sources as a gathering point, striking evidence of the permanence of the strategic role played by the Abbad area.186
86It shows that the Ptolemaic logistical network was not sufficient to rely on and that the system was also based on the supply of caravans at the beginning of the road, with a sufficient quantity of water to cross the first part of the desert and reach a hydreuma where they could be supplied again.187 The reconstitution of the Ptolemaic network is just at its beginning188 but the ostraca found at Bir Samut show, for example, that the system relied on four main stops, including the fort near the Paneion of al‑Kanais, the site of Pherembesis, that could be Bir Samut, and two yet to be localized sites, the “convenient well of Arsinoe” (τὸ Ἀρcινόηc Εὔκαιρον ὕδρευμα),189 and “the supplementary well” (Προcεξευρέθεν ὕδρευμα).190 This last name is significant, as it appears to suggest that the Ptolemaic administration realized, after the road was first equipped with three supply stops, that travelers needed an additional stop to ensure their journey ran smoothly and with reduced uncertainty. It is also noticeable that smaller sites – likely mining sites in the vicinity of the forts and stations – may have been used to supply water to the main stops, as shown by the “ostracon of the miracle” already discussed.
87Wells were therefore rare, at least at the beginning of the equipment of the Edfu‑Berenike road. Moreover, the size of the Ptolemaic cisterns was remarkably small compared to the cisterns of Roman forts.191 The sāqia was not invented at that time and people could only rely on shadufs to fill them with water. By analogy with the shift that occurred in the 1830s‑1850s in the Eastern Desert, this leads us to wonder whether in the Ptolemaic period, when the network was not as coordinated as in Roman times, the travelers were more empirical in getting water supplies. One should note for example that one stop on the hypothetical Ptolemaic road leading from Coptos to Berenike,192 the Paneion of the Wadi Menih al‑Hayr, is situated at 3 km only from a temporary lake (where the so-called ḥafr of the Wadi Menih al‑Hayr is located).193 The stop had no built structures and caravans and expeditions likely slept one night here, to rest in the desert. The site can be identified as the first “in monte” stop listed by Pliny (see above), and although the travelers had no well at their disposal, the temporary lake was surely known to them. At some point, a wall was built, probably to protect a well inside the lake, in order to secure the water availability.194 In Pliny’s list, another “in monte” stop is mentioned between Daghbag/Compasi and Apollonos hydreuma. Unfortunately, this stop has not been located but there is a good chance that this stop was in an area where water was readily available in the form of a spring or a qalt, or by digging a few decimeters into the wadi floor195 to reach a locally high water table in the wadi alluvium or to find water in a sediment infilled qalt.196
88Finally, the last point we would like to discuss is the possible role of nomads in the management of resources in Antiquity. Their Bedouin guides’ knowledge of wells was a crucial factor in the smooth running of journeys, a fact repeatedly mentioned in accounts by travelers. No source is clear enough to postulate it for Antiquity, although for the Ptolemaic period one must insist on the fact that an unknown number of guides (ὁδηγοὶ) were part of Lichas’ expedition documented by the Abbad ostraca.197 They certainly could have guided the elephant hunters on African soil, after the transportation of the troops by boat from Berenike. However, they may also have had a role in the Eastern Desert itself, notably to ensure the security of the water supply. Another guide, with a foreign name, is mentioned acting on the same road according to H. Cuvigny, in O.Sam. inv. 1275, included in one unit of the royal camel drivers who roamed the desert roads in the second half of the 3rd century BC.198 This person probably only worked in the Eastern Desert, which reinforces our hypothesis. The integration of the local population is not unattested for the Ptolemaic period and, in another strategic field of road logistics, the Ptolemies did not hesitate to rely on the “Arabs of the desert”, a local population around Bir Samut, to raise the camels they used for caravans to Berenike.199 Local people may also have been integrated in food logistics too, as shown by the discovery in significant quantities, in some kitchens of Bir Samut, of locally produced cooking pots with patterns reminiscent of the well-known ceramics of the desert dwellers of Roman times, the Blemmyes.200 They also appear in the ostraca of Bir Samut as beneficiaries of food distribution, like other people traveling on the road, and thus seem to be partially integrated into the Ptolemaic road system, and water management may have been one of their areas of expertise.201
Conclusion: water, connectivity and networks in Antiquity and modern times
89The Eastern Desert of Egypt, which might seem a priori an obstacle to passage, actually constitutes an interface between the Nile Valley, in contact with the Mediterranean and African worlds, and a seafront open to the Red Sea and the Indian Ocean. This helps to explain why, in spite of its desert environment and extreme aridity, it has not been deserted by humans and has been rather densely occupied in comparison with other deserts in the world. Incidentally, water is available in the Eastern Desert in various forms as long as one knows how to find the water resources and maintain the water equipment. To do so, keeping in memory the location of water points was critical at all times, and could certainly only have been done with local knowledge and an efficient administrative and logistical management during the periods when the desert was controlled from the Nile Valley.
90The water resource, its quantity and quality, was therefore at the center of concerns of people and states, from the expeditions to Punt and Sinai in the 3rd millennium BC to the last commercial crossings during the 20th century AD. The importance of the Eastern Desert has been somewhat forgotten since the increase in the number of motorized boats and the digging of the Suez Canal, which connects the Red Sea directly to the Mediterranean Sea without passing overland. But before that, as the Red Sea was difficult to navigate in its northern part, the Eastern Desert corresponded to a safer and more practical passage.202 Its importance becomes clearer when we refer to the French expedition in Egypt (1798‑1801): the occupation of Laqeita and Quseir was enough to temporarily stop reinforcements coming from Mecca to support the Mamelukes, and the occupation of Egypt in general disrupted the organization of trade in the British Empire.203
91Within this desert, integrated into global networks, several distinct networks coexist and/or overlap according to different objectives:204
- travelers, merchants and pilgrims (in modern times) who only seek to cross the desert (or to transport goods) use extra-regional networks and visit main road nodes to connect the poles of the Nile Valley to those of the Red Sea coast;
- nomads and workers exploiting the resources, especially mineral resources, of the desert have to ensure their subsistence locally, they practice intra-regional networks and are more likely to visit secondary nodes.
92Obviously, these networks are not separated and some nodes related to water resources are common, such as Laqeita, Quseir and some well supplied wells in Rawd al‑Buram or on the road from Qena to Quseir in modern times. In Antiquity, Laqeita/Phoinikon, Daghbag/Compasi, Bir Samut, al‑Hammamat and many others have played such a role. It thus appears that the integration of nodes related to water resources in the networks of the Eastern Desert, and more broadly those of the Red Sea region, is variable. In addition to the major nodes mentioned above, secondary nodes linked to the major sites are attested, for example the water points of the Wadi Tarfawi area, which are not strictly speaking integrated into the road network, but are strongly connected to Quseir. In Antiquity, the water supply of the two harbors of Myos Hormos and Berenike, which has not been discussed here, shared the same characteristic, as well as the hydraulic system evidenced in the quarrying areas of Mons Claudianus and the Porphyrites.
93The integration of these networks into a macro-regional system has evolved over time as a result of international relations or economic or technological changes, thus influencing the destiny of their internal nodes. For example, Quseir had a brilliant period under Mohamed Ali, after more difficult times in the 18th and early 19th century, before losing importance between 1850 and 1900 due to multiple factors (construction of a railroad from the Nile to Suez, growth in the number of steamboats making maritime transport safer and more regular, construction of the Suez Canal, competition for wheat from the Tigris and Euphrates rivers to supply Mecca).205 This marginalization of Quseir and of the road networks of the Eastern Desert had drastic consequences on its secondary nodes and many wells have now totally collapsed and/or are infilled. New ones have been dug, in new locations, that are more convenient for the supply of the present-day network of roads and settlements meant to function at an intra-national scale, with few or no connection to the Red Sea. It recalls the situation of ancient Myos Hormos, located 7 km north of Quseir, whose foundation in the Ptolemaic period was at the origin of the development of many secondary nodes in its vicinity, such as Pliny’s Tadnos Fons, or the sites of Bir Karim,206 and Bir Nakheil to supply water to the city (which had no local sweet water source). Its abandonment during the 2nd century AD, due to the silting up of its port, and perhaps the reorientation of Indian trade, is at the origin of the gradual abandonment of the hydraulic system of the road from Coptos, at the same time the Roman forts were abandoned. The hydraulic network though was partly reactivated when Quseir, which took the place of Quseir al‑Qadim during medieval times, experienced a last revival under Mohamed Ali.
94However, by making all these comparisons between modern and ancient times, we do not want to end here with the thought that history repeats itself. Each period obviously has its peculiarities, but it is worth pointing out one striking feature: the stability of the location and function of certain nodes over time, on the Red Sea shore and in the Valley, and also in the desert itself, the latter ones in close relation to the availability of water they gave access to.
Bibliographie
Abbreviations
Papyrological editions are cited in accordance with the Checklist of editions of Greek, Latin, demotic, and Coptic papyri, ostraca, and tablets (http://papyri.info/docs/checklist).
BAEFE: Bulletin archéologique des écoles françaises à l’étranger (s.l.).
BIFAO: Bulletin de l’Institut français d’archéologie orientale (Cairo).
BSGE: Bulletin de la Société de géographie d’Égypte (Cairo).
ILS: H. Dessau (ed.), Inscriptiones Latinae selectae, 3 vol., Berlin, Berolini, 1892‑1916.
I.Paneion = Bernand 1972.
I.Pan = Bernand 1977.
O.Ber. II = Bagnall, Helms, Verhoogt 2005.
O.Ber. III = Ast, Bagnall 2016.
Works cited
Agut‑Labordère 2018: D. Agut‑Labordère, “L’introduction du dromadaire dans le désert occidental égyptien au Ier millénaire av. J.‑C.”, in G. Tallet, T. Sauzeau (ed.), Mers et déserts de l’Antiquité à nos jours: approches croisées, Rennes, Presses universitaires de Rennes, 2018, pp. 175‑195.
Agut‑Labordère, Redon 2020: D. Agut‑Labordère, B. Redon (ed.), Les vaisseaux du désert et des steppes. Les camélidés dans l’Antiquité (Camelus dromedarius et Camelus bactrianus), Lyon, MOM Éditions, 2020, https://books.openedition.org/momeditions/8457 (accessed 04/06/2020).
Aldsworth, Barnard 1996: F.G. Aldsworth, H. Barnard, “Survey of Hitan Rayan”, in S.E. Sidebotham, W. Wendrich (ed.), Berenike 1995: preliminary report of the 1995 excavations at Berenike (Egyptian Red Sea coast) and the survey of the Eastern Desert, Leiden, Research School CNWS, 1996, pp. 411‑440.
Ast 2018: R. Ast, “Berenike in light of inscriptions, ostraca, and papyri”, in J.‑P. Brun, T. Faucher, B. Redon, S. Sidebotham (ed.), The Eastern Desert of Egypt during the Greco-Roman Period: archaeological reports, Paris, Collège de France, 2018, https://books.openedition.org/cdf/5232 (accessed 15/07/2020).
Ast, Bagnall 2016: R. Ast, R.S. Bagnall, O.Berenike III. Greek and Latin texts from the 2009‑2013 seasons, Pap.Brux. XXXVI, Brussels, Association égyptologique Reine Élisabeth, 2016.
d’Athanasi 1836: G. d’Athanasi, A brief account of the researches and discoveries in Upper Egypt made under the direction of H. Salt esq., by Giovanni d’Athanasi, to which is added a detailed catalogue of Mr Salt’s collection of Egyptian antiquities, London, John Hearne, 1836, https://archive.org/details/abriefaccountre00athagoog (accessed 15/01/2021).
Bagnall, Bülow‑Jacobsen, Cuvigny 2001: R.S. Bagnall, A. Bülow‑Jacobsen, H. Cuvigny, “Security and water on Egypt’s desert roads: new light on the Prefect Iulius Ursus and praesidia-building under Vespasian”, JRA 14, 2001, pp. 325‑333.
Bagnall, Helms, Verhoogt 2005: R.S. Bagnall, C. Helms, A.M.F.W. Verhoogt, O.Berenike II. Texts from the 1999‑2001 seasons, Pap.Brux. XXXIII, Brussels, Association égyptologique Reine Élisabeth, 2005.
Ball 1912: J. Ball, The geography and geology of south-eastern Egypt, Cairo, Government Press, 1912.
Ball 1927: J. Ball, “Problems of the Libyan Desert (continued)”, The Geographical Journal 70/2, 1927, pp. 105‑128.
Barron, Hume 1902: T. Barron, W.F. Hume, The topography and geology of the Eastern desert of Egypt: central portion, Cairo, National Printing Department, 1902, https://archive.org/details/topographyandge01humegoog (accessed 15/01/2021).
Belzoni 1820: G.B. Belzoni, Narrative of the operations and recent discoveries within the pyramids, temples, tombs and excavations in Egypt and Nubia, and of a journey to the coast of the Red Sea, in search of the ancient Berenice and another to the oasis of Jupiter Ammon, London, John Murray, 1820, https://gallica.bnf.fr/ark:/12148/bpt6k1053464 (accessed 04/07/2018).
Bernand 1972: A. Bernand, Le Paneion d’el‑Kanaïs. Les inscriptions grecques, Leiden, Brill, 1972.
Bernand 1977: A. Bernand, Pan du désert, Leiden, Brill, 1977.
Bisson de la Roque 1922: F. Bisson de la Roque, “Voyage au Djebel Shaïb”, BSGE 11/3‑4, 1922, pp. 113‑140.
Bonneau 1993: Bonneau, Le régime administratif de l’eau du Nil dans l’Égypte grecque, romaine et byzantine, Leiden/New York/Cologne, Brill, 1993.
Browne 1799: W.G. Browne, Travels in Africa, Egypt, and Syria, from the year 1792 to 1798, London, Paternoster-Row, 1799 (printed for T. Cadell junior and W. Davies, Strand and T.N. Longman and O. Rees).
Bruce 1790: J. Bruce, Travels to discover the source of the Nile, in the years 1768, 1769, 1770, 1771, 1772, and 1773, Londres/Edinburgh, printed by J. Ruthven, for G.G.J. and J. Robinson, London, 1790, https://archive.org/details/travelstodiscove01bruc (accessed 04/07/2018).
Brun 2018: J.‑P. Brun, “Chronology of the forts of the routes to Myos Hormos and Berenike during the Graeco-Roman period”, in J.‑P. Brun, T. Faucher, B. Redon, S.E. Sidebotham (ed.), The Eastern Desert of Egypt during the Greco-Roman period: archaeological reports, Paris, Collège de France, 2018, https://books.openedition.org/cdf/5239 (accessed 15/07/2020).
Brun et al. 2018: J.‑P. Brun, T. Faucher, B. Redon, S. Sidebotham (ed.), Le désert oriental d’Égypte durant la période gréco-romaine: bilans archéologiques, Paris, Collège de France, 2018, https://books.openedition.org/cdf/4932 (accessed 10/04/2018).
Bubenzer, Riemer 2007: O. Bubenzer, H. Riemer, “Holocene climatic change and human settlement between the central Sahara and the Nile Valley – Archaeological and geomorphological results”, Geoarchaeology 22/6, 2007, pp. 607‑620.
Bülow‑Jacobsen 2003: A. Bülow‑Jacobsen, “Toponyms and proskynemata”, in H. Cuvigny (ed.), La route de Myos Hormos. L’armée romaine dans le désert Oriental d’Egypte, Cairo, IFAO, 2003, pp. 51‑59.
Bülow‑Jacobsen, Cuvigny 2007: A. Bülow‑Jacobsen, H. Cuvigny, “Sulpicius Serenus, procurator Augusti, et la titulature des préfets de Berenice”, Chiron 37, 2007, pp. 11‑33.
Bülow‑Jacobsen, Cuvigny, Fournet 1994: A. Bülow‑Jacobsen, H. Cuvigny, J.‑L. Fournet, “The identification of Myos Hormos. New papyrological evidence”, BIFAO 94, 1994, pp. 27‑42.
Cailliaud 1821: F. Cailliaud, Voyage à l’oasis de Thèbes et dans le déserts situés à l’orient et à l’occident de la Thébaïde, fait pendent les années 1815, 1816, 1817 et 1818, Paris, Imprimerie royale, 1821, https://gallica.bnf.fr/ark:/12148/bpt6k106154c.image (accessed 15/01/2021).
Chaufray 2020: M.‑P. Chaufray, “Les chameaux dans les ostraca démotiques de Bi’r Samut (Égypte, désert Oriental)”, in B. Redon, D. Agut-Labordère (ed.), Les vaisseaux du désert et des steppes. Les camélidés dans l’Antiquité (Camelus dromedarius et Camelus bactrianus), Lyon, MOM Éditions, 2020, pp. 135‑170, https://books.openedition.org/momeditions/8567 (accessed 04/06/2020).
Chaufray forthcoming: M.‑P. Chaufray, “Blemmyes. New documents and new perspectives”, in H. Cuvigny (ed.), Blemmyes. New documents and new perspectives, Cairo, IFAO, forthcoming.
Chaufray, Redon 2021: M.‑P. Chaufray, B. Redon, “Ostraca and tituli picti of Samut North and Bi’r Samut (Eastern Desert of Egypt). Some reflections on find location”, in J. Lougovaya, C. Caputo (ed.), Using ostraca in the Ancient World. New discoveries and methodologies, Berlin/Boston, De Gruyter, 2021, pp. 163‑180.
Colston 1886: R.E. Colston, “Journal d’un voyage du Caire à Keneh, Bérénike et Berber, et retour par le désert de Korosko (1873‑1874)”, BSGE 2/9, 1886, pp. 489‑568.
Cooper 2011: J.P. Cooper, “No easy option: Nile versus Red Sea in ancient and medieval north-south navigation”, in W.V. Harris, K. Iara (ed.), Maritime technology in the ancient economy: ship-design and navigation, published in Journal of Roman Archaeology Supplement 84, 2011, pp. 189‑210.
Couyat 1909: J. Couyat, “Alexis Bert. Description du désert de Siout à la mer Rouge, d’après un manuscrit de la Bibliothèque royale de Turin. Relation d’une course faite pour reconnaître une partie du désert et des montagnes à l’est de Siouth”, BIFAO 9, 1909, pp. 139‑184.
Couyat 1910: J. Couyat, “Alexis Bert. Description du désert de Siout à la mer Rouge, d’après un manuscrit de la Bibliothèque royale de Turin. Relation d’une course faite pour reconnaître une partie du désert et des montagnes à l’est de Siouth (suite)”, BIFAO 10, 1910, pp. 1‑77.
Crépy, Manière, Redon forthcoming: M. Crépy, L. Manière, B. Redon, “Roads in the sand: using data from modern travelers to reconstruct the ancient road networks of Egypt’s Eastern Desert”, in T. Kulayci (ed.), Archaeologies of roads, s.l., Digital Press at the University of North Dakota, forthcoming.
Crépy, Redon 2020: M. Crépy, B. Redon, “Début et fin de l’usage intensif des chameaux dans le désert Oriental d’Égypte”, Ethnozootechnie 106, 2020, pp. 21‑28.
Cuvigny 1997: H. Cuvigny, “Le crépuscule d’un dieu: le déclin du culte de Pan dans le désert Oriental”, BIFAO 97, 1997, pp. 139‑147.
Cuvigny 2003a: H. Cuvigny, “Les documents écrits de la route de Myos Hormos à l’époque gréco-romaine”, in H. Cuvigny (ed.), La route de Myos Hormos. L’armée romaine dans le désert Oriental d’Égypte, Cairo, IFAO, 2003, pp. 265‑294.
Cuvigny 2003b: H. Cuvigny, “Le fonctionnement du réseau”, in H. Cuvigny (ed.), La route de Myos Hormos. L’armée romaine dans le désert Oriental d’Égypte, Cairo, IFAO, pp. 295‑359.
Cuvigny 2003c: H. Cuvigny, “La société civile des praesidia”, in H. Cuvigny (ed.), La route de Myos Hormos. L’armée romaine dans le désert Oriental d’Égypte, Cairo, IFAO, 2003, pp. 361‑398.
Cuvigny 2003d: H. Cuvigny (ed.), La route de Myos Hormos. L’armée romaine dans le désert Oriental d’Égypte, Cairo, IFAO, 2003.
Cuvigny 2005: H. Cuvigny, Ostraca de Krokodilô. La correspondance militaire et sa circulation, Cairo, IFAO, 2005.
Cuvigny 2010: H. Cuvigny, “The shrine in the praesidium of Dios (Eastern Desert of Egypt): graffiti and oracles in context”, Chiron 40, 2010, pp. 245‑299.
Cuvigny 2011: H. Cuvigny (ed.), Didymoi. Une garnison romaine dans le désert Oriental d’Égypte I. Les fouilles et le matériel, Cairo, IFAO, 2011.
Cuvigny 2012: H. Cuvigny (ed.), Didymoi. Une garnison romaine dans le désert Oriental d’Égypte II. Les textes, Cairo, IFAO, 2012.
Cuvigny 2017: H. Cuvigny, “Quand Lichas plantait sa tente à Abbad. Un dossier de distribution d’eau sur la route d’Edfou à Bérénice (c. 240‑210a)”, CE 92/183, 2017, pp. 111‑128.
Cuvigny 2018: H. Cuvigny, “A survey of place-names in the Egyptian Eastern Desert during the principate according to the ostraca and the inscriptions”, in J.‑P. Brun, T. Faucher, B. Redon, S. Sidebotham (ed.), The Eastern Desert of Egypt during the Greco-Roman period: archaeological reports, Paris, Collège de France, 2018, https://books.openedition.org/cdf/5231 (accessed 15/07/2020).
Cuvigny 2020: H. Cuvigny, “L’élevage des chameaux sur la route d’Edfou à Bérénice d’après une lettre trouvée à Bi’r Samut (iiie siècle av. J.‑C.)”, in B. Redon, D. Agut‑Labordère (ed.), Les vaisseaux du désert et des steppes. Les camélidés dans l’Antiquité (Camelus dromedarius et Camelus bactrianus), Lyon, MOM Éditions, 2020, pp. 171‑180, https://books.openedition.org/momeditions/8572 (accessed 04/06/2020).
Cuvigny et al. 2010: H. Cuvigny, E. Botte, J.‑P. Brun, A. Bülow‑Jacobsen, M. Reddé, “Xeron 2010”, unpublished report, https://www.igl.ku.dk/bulow/Xeron10.pdf (accessed 15/01/2021).
De Noé 1826: L.P.J.A. De Noé, Mémoires relatifs à l’expédition anglaise partie du Bengale, en 1800, pour aller combattre en Égypte l’armée d’Orient, s.l., Imprimerie royale, 1826.
De Romanis 1996: F. De Romanis, Cassia, cinnamomo, ossidiana. Uomini e merci tra Oceano Indiano e Mediterraneo, Rome, “L’Erma” di Bretschneider, 1996.
De Rozière 1801‑1802: F.M. De Rozière, “Description minéralogique de la vallée de Qosséyr, lue à l’institut d’Égypte, dans les séances des 21 brumaire au 11 frimaire de l’an 8”, Journal des Mines 66. Ventôse, 1801‑1802, pp. 449‑488, http://annales.ensmp.fr/articles/1801-1802-1/245-265.pdf (accessed 15/01/2021).
De Rozière 1813: F.M. De Rozière, “Description minéralogique de la Vallée de Qoceyr”, in E.‑F. Jomard (ed.), Description de l’Égypte: histoire naturelle, vol. II, Paris, Imprimerie impériale, 1813, pp. 86‑98, https://digi.ub.uni-heidelberg.de/diglit/jomard1813bd5_1_2/0001/image (accessed 15/01/2021).
Denon 1802: V. Denon, Voyage dans la Basse et la Haute Égypte, pendant les campagnes du général Bonaparte, Paris, Impr. P. Didot l’aîné, 1802, https://gallica.bnf.fr/ark:/12148/bpt6k5787505v (accessed 03/06/2020).
Desanges 2008: Pline l’Ancien, Histoire naturelle, livre VI, ed. and transl. J. Desanges, Collection des Universités de France, Paris, Les Belles Lettres, 2008.
Du Camp 1860: M. Du Camp, Le Nil: Égypte et Nubie, Paris, A. Bourdilliat et Cie, 1860 (2nd ed.), https://gallica.bnf.fr/ark:/12148/bpt6k5609167r (accessed 04/07/2018).
Dubois‑Aymé 1802: J.‑M. Dubois‑Aymé, “Notice sur Qosséir et ses environs”, in Mémoires sur l’Égypte, publiés dans les années VII, VIII et IX, t. 3, Paris, Impr. P. Didot l’aîné, 1802 (An X), pp. 273‑285, https://babel.hathitrust.org/cgi/pt?id=mdp.39015014669421&view=2up&seq=284 (accessed 15/01/2021).
Embabi 2004: N.S. Embabi, The geomorphology of Egypt: landforms and evolution, Cairo, Egyptian Geographical Society, 2004 (1st ed.).
Ezzat 1974: M.A. Ezzat, Groundwater series in the Arab Republic of Egypt; exploitation of groundwater in El Wadi El Gedid project area. General desert development, Ministry of Irrigation (Egypt), part I‑IV, Cairo, Ministry of Irrigation, 1974.
Faucher et al. 2020: T. Faucher, B. Redon, M. Crépy, A. Rabot, J. Gates‑Foster, C. Bouchaud, V. Dabrowski, N. Villars, “Désert Oriental – Époque ptolémaïque”, BAEFE. Égypte 2020, 2020, https://journals.openedition.org/baefe/1087 (accessed 08/09/2020).
Flaubert 1910: G. Flaubert, Œuvres complètes de Gustave Flaubert. Notes de voyages I. Italie – Égypte – Palestine – Rhodes, Paris, Conard libraire-éditeur, 1910.
Floyer 1887: E.A. Floyer, “Notes on a sketch map of two routes in the Eastern Desert of Egypt”, Proceedings of the Royal Geographical Society and Monthly Record of Geography 9/11, 1887, pp. 659‑681.
Floyer 1893: E.A. Floyer, Étude sur le nord-Etbai entre le Nil et la mer Rouge, Cairo, Impr. nationale, 1893.
Gasse 2000: F. Gasse, “Hydrological changes in the African tropics since the last glacial maximum”, Quaternary Science Reviews 19/1‑5, 2000, pp. 189‑211.
Golénischeff 1890: W.S. Golénischeff, “Une excursion à Bérénice”, Recueil de travaux relatifs à la philologie et à l’archéologie égyptiennes et assyriennes 13, 1890, pp. 75‑96, https://digi.ub.uni-heidelberg.de/diglit/rectrav1890/0086 (accessed 19/01/2021).
Guimier‑Sorbets, Seif El‑Din 1997: A.‑M. Guimier‑Sorbets, M. Seif El‑Din, “Les deux tombes de Perséphone dans la nécropole de Kom el‑Chougafa à Alexandrie”, BCH 121/1, 1997, pp. 355‑410.
Hinkel 1991: M. Hinkel, “Hafire im antiken Sudan”, ZÄS 118/1, 1991, pp. 32‑47.
Hobbs 2014: J.J. Hobbs, “Bedouin place names in the Eastern Desert of Egypt”, Nomadic Peoples 18/2. Reshaping tribal identities in the contemporary Arab world, 2014, pp. 123‑146.
Jones 1932: Strabo, Geography, volume VIII, book 17, transl. H.L. Jones, Loeb Classical Library 267, Cambridge, MA, Harvard University Press, 1932.
Klemm, Klemm 2013: D.D. Klemm, R. Klemm, Gold and gold mining in Ancient Egypt and Nubia: geoarchaeology of the ancient gold mining sites in the Egyptian and Sudanese Eastern Deserts, Berlin, Springer, 2013.
Klunzinger 1878: C.B. Klunzinger, Upper Egypt: its people and its products: a descriptive account of the manners, customs, superstitions, and occupations of the people of the Nile Valley, the desert, and the Red Sea coast; with sketches of the natural history and geology, New York, Scribner, Armstrong & Co, 1878, https://archive.org/details/upperegyptitspeo00klunrich (accessed 04/07/2018).
Krzywinski, Seland 2007: J. Krzywinski, E.H. Seland, “Water harvesting in the Eastern Desert of Egypt”, in E.H. Seland (ed.), The Indian Ocean in the Ancient Period. Definite places, translocal exchange, Oxford, Archaeopress, 2007, pp. 45‑57.
L’Hôte 1841: N. L’Hôte, “Lettres sur l’Égypte en 1841. Quosseyr – Les mines d’émeraude”, Revue des deux mondes – quatrième série 27/1, 1841, pp. 136‑151, https://www.jstor.org/stable/44689315 (accessed 19/01/2021).
Lichtheim 1976: M. Lichtheim, Ancient Egyptian literature II. The New Kingdom, Berkeley/Los Angeles/London, University of California Press, 1976 (1st ed.).
Linant de Bellefonds 1868: L.M.A. Linant de Bellefonds, L’Étbaye, pays habité par les Arabes Bicharieh. Géographie, ethnographie, mines d’or, Paris, Arthus Bertrand, 1868, https://gallica.bnf.fr/ark:/12148/bpt6k9106456p (accessed 04/07/2018).
Mac Alister 1900: D.A. Mac Alister, “The emerald mines of northern Etbai”, The Geographical Journal 16/5, 1900, pp. 537‑549.
Manière, Crépy, Redon 2021: L. Manière, M. Crépy, B. Redon, “Building a Model to Reconstruct the Hellenistic and Roman Road Networks of the Eastern Desert of Egypt, a Semi-Empirical Approach Based on Modern Travelers’ Itineraries”, Journal of Computer Applications in Archaeology, 2021, pp. 20‑46.
Maxfield, Peacock 2001: V.A. Maxfield, D.P.S. Peacock, The Roman imperial quarries. Survey and excavation at Mons Porphyrites 1994‑1998 I. Topography and quarries, London, Egypt Exploration Society, 2001.
Murray 1925: G.W.W. Murray, “The Roman roads and stations in the Eastern Desert of Egypt”, JEA 11/3‑4, 1925, pp. 138‑150.
Murray 1952: G.W.W. Murray, The artesian water of Egypt, Cairo, Ministry of Finance Survey of Egypt, 1952.
Murray 1955: G.W.W. Murray, “Water from the desert: some ancient Egyptian achievements”, The Geographical Journal 121/2, 1955, pp. 171‑181.
Prickett 1979: M. Prickett, “Quseir regional survey”, in D.S. Whitcomb, J.H. Johnson (ed.), Quseir al‑Qadim 1978. Preliminary report, Cairo/New Jersey, American Research Center in Egypt, 1979, pp. 257‑352.
Raimondi 1923: J. Raimondi, Le désert Oriental égyptien: du Nil à la Mer Rouge; ses richesses dans le passé, son importance dans l’avenir, Cairo, Impr. de l’IFAO pour la Société royale de géographie d’Égypte, 1923.
Reddé 2018: M. Reddé, “The fortlets of the Eastern Desert of Egypt”, in J.‑P. Brun, T. Faucher, B. Redon, S.E. Sidebotham (ed.), The Eastern Desert of Egypt during the Greco-Roman period: archaeological reports, Paris, Collège de France, 2018, https://books.openedition.org/cdf/5248 (accessed 15/07/2020).
Redon 2018: B. Redon, “The Control of the Eastern Desert by the Ptolemies: new archaeological data”, in J.‑P. Brun, T. Faucher, S.E. Sidebotham (ed.), The Eastern Desert of Egypt during the Greco-Roman period: archaeological reports, Paris, Collège de France, 2018, https://books.openedition.org/cdf/5249 (accessed 15/07/2020).
Redon, Faucher 2020: B. Redon, T. Faucher, Samut Nord. L’exploitation de l’or du désert Oriental à l’époque ptolémaïque, Cairo, IFAO, 2020.
Sanlaville 1997: P. Sanlaville, “Les changements dans l’environnement au Moyen-Orient de 20 000 BP à 6 000 BP”, Paléorient 23/2, 1997, pp. 249‑262.
Scaife 1935: C.H.O. Scaife, Two inscriptions at Mons Porphyrites (Gebel Dokhan). Also a description, with plans, of the stations between Kainopolis & Myos Hormos together with some other ruins in the neighborhood of Gebel Dokhan, Cairo, University of Cairo, 1935.
Sidebotham 1989: S.E. Sidebotham, “Ports of the Red Sea and the Arabia-India trade”, in D.H. French, C.S. Lightfoot (ed.), The eastern frontier of the Roman Empire. Proceedings of a colloquium held at Ankara in September 1988, Oxford, British Archaeological Reports, 1989, pp. 485‑513.
Sidebotham 1994: S.E. Sidebotham, “University of Delaware fieldwork in the Eastern Desert of Egypt, 1993”, DOP 48, 1994, pp. 263‑275.
Sidebotham 2011: S.E. Sidebotham, Berenike and the ancient maritime spice route, Berkeley, University of California Press, 2011.
Sidebotham, Gates‑Foster 2019: S.E. Sidebotham, J.E. Gates‑Foster (ed.), The archaeological survey of the desert roads between Berenike and the Nile Valley. Expeditions by the University of Michigan and the University of Delaware to the Eastern Desert of Egypt, 1987‑2015, Boston, American Schools of Oriental Research, 2019.
Sidebotham, Hense, Nouwens 2008: S.E. Sidebotham, M. Hense, H.M. Nouwens, The Red Land. The illustrated archaeology of Egypt’s Eastern Desert, Cairo, American University in Cairo Press, 2008.
Sidebotham et al. 2004: S.E. Sidebotham, G.T. Mikhail, J.A. Harrell, R.S. Bagnall, “A water temple at Bir Abu Safa (Eastern Desert, Egypt)”, JARCE 41, 2004, pp. 149‑159.
Tregenza 1955: L.A. Tregenza, The Red Sea mountains of Egypt, London, Oxford University Press, 1955.
Van der Veen et al. 2018: M. Van der Veen, C. Bouchaud, R. Cappers, C. Newton, “Roman life in the Eastern Desert of Egypt: food, imperial power and geopolitics”, in J.‑P. Brun, T. Faucher, B. Redon, S.E. Sidebotham (ed.), The Eastern Desert of Egypt during the Greco-Roman period: archaeological reports, Paris, Collège de France, 2018, https://books.openedition.org/cdf/5252 (accessed 15/07/2020).
Van Rengen 2018: W. Van Rengen, “Some topographical problems around Myos Hormos: Philotera‑Philoteris”, in J.‑P. Brun, T. Faucher, B. Redon, S.E. Sidebotham (ed.), The Eastern Desert of Egypt during the Greco-Roman period: archaeological reports, Paris, Collège de France, 2018, https://books.openedition.org/cdf/5253 (accessed 15/07/2020).
Weigall 1913: A.E.P.B. Weigall, Travels in the Upper Egyptian deserts, Edinburgh/London, Blackwood, 1913.
Weschenfelder 2012: P. Weschenfelder, “Water management by divine benevolence along the Nile River: artificial water reservoirs as pastoral meeting places in the Meroitic Sudan (ca. 350 B.C. to 350 A.D.)”, in H.P. Hahn, K. Cless, J. Soentgen (ed.), People at the well. Kinds, usages and meanings of water in a global perspective, Frankfurt am Main/New York, Campus Verlag, 2012, pp. 247‑265.
Wilkinson 1832: J.G. Wilkinson, “Notes on a part of the Eastern Desert of Upper Egypt”, The Journal of the Royal Geographical Society of London 2, 1832, pp. 28‑60.
Wilkinson 1847: J.G. Wilkinson, Handbook for travellers in Egypt: including descriptions of the course of the Nile to the second cataract, Alexandria, Cairo, the pyramids, and Thebes, the overland transit to India, the peninsula of Mount Sinai, the oases, & c., London, John Murray, 1847.
Williams 2018: M. Williams, The Nile basin: quaternary geology, geomorphology and prehistoric environments, s.l., Cambridge University Press, 2018 (1st ed.).
Notes de bas de page
1 Sanlaville 1997, p. 259.
2 On the expeditions through the desert and then the Red Sea, see the article by C. Somaglino and P. Tallet in this volume. On the road networks of the region in the Middle and New Kingdom, see the article by I. Goncalves. For the Greco-Roman period, very good generalist presentations of the region and its history can be found in Sidebotham, Hense, Nouwens 2008 and Sidebotham 2011.
3 The bibliography is too extensive to be cited here. See recently Brun et al. 2018; Klemm, Klemm 2013; Sidebotham, Gates‑Foster 2019.
4 This project, led by B. Redon, has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (grant agreement no. 759078).
5 See Crépy, Redon 2020.
6 For example, Sidebotham, Gates‑Foster 2019; Sidebotham, Hense, Nouwens 2008. These data have been updated recently by the Desert Networks team, and published on the following website: https://desertnetworks.huma-num.fr/. It presents a gazetteer of more than 260 archaeological sites and more than 200 watering places located in the Eastern desert, and several interactive maps.
7 For example, Cuvigny 2018.
8 Following the same approach, it was also possible to trace some of the routes used by travelers between the 18th and 20th centuries and to compare them with ancient routes, in order to identify continuities and breaks: see Manière, Crépy, Redon 2021 and Crépy, Manière, Redon forthcoming.
9 Due to lack of space in this article, we will not discuss the particular case of ancient and modern ports (Myos Hormos/Quseir al‑Qadim and Berenike in Greco-Roman times, Quseir in modern times) which do not have their own water resources and therefore rely on the water network of the desert.
10 Data from the New Kingdom will be mentioned here when possible. See the article by I. Goncalves in this volume for a more focused study of this period.
11 The MAFDO, founded by H. Cuvigny (CNRS, IRHT) in 1994, and directed by B. Redon from 2013 to 2017 is currently led by T. Faucher (CNRS, IRAMAT-CRP2A). Since 2013, five sites have been excavated: Samut North, a gold-mining settlement, Bir Samut and Abbad, located along the road leading from Edfu to Berenike, Deir el-Atrash, a Roman fort on the road to the Porphyrites quarries, and Ghozza, a mining settlement in the northern desert region, an area of quarries and mines.
12 See, among others, Cuvigny 2003d; Cuvigny 2005; Cuvigny 2011; Cuvigny 2012.
13 Except for very short lengths directly linked to permanent springs such as Ayn al‑Ghazal, in the vicinity of Quseir.
14 Bubenzer, Riemer 2007, pp. 610‑611.
15 Sanlaville 1997, p. 259.
16 Embabi 2004, p. 13.
17 Ezzat 1974.
18 This is why tanks – although largely used in Antiquity – are not discussed here, since the travelers never mention them in their accounts.
19 The water points were recorded and georeferenced from the maps by A. Rabot (HiSoMA, university Lyon 2), L. Manière (HiSoMA, ERC Desert Networks project) and H. Verreth (ERC Desert Networks project and Trismegistos). See https://desertnetworks.huma-num.fr/.
20 A spring called Amusne described by travelers.
21 A qalt (pl. qilāt) is a natural surface basin filled with rain water and/or spring water, see infra. We thank the anonymous reviewer of this article and Naïm Vanthieghem for their kind help in the transliteration of the Arabic words.
22 One of the main qualities of a qalt is that it is in the shade most of the day, so it is frequently hidden on satellite images.
23 The main road used to cross the desert in the Ptolemaic period was the road from Edfu to Berenike. In the Roman period, the two major roads start from Coptos on the Nile Valley, reaching Myos Hormos and Berenike on the Red Sea shore (see fig. 1).
24 The two sites are only 7 km apart.
25 For more information about this road, see Cuvigny 2003d.
26 Approximately 20 km separates these two localities. After Weigall 1913, p. 64, during medieval times, Qus was the valley terminus of this road.
27 Wilkinson 1847, pp. 398‑399; Floyer 1887, p. 662.
28 Browne 1799, pp. 146‑147.
29 De Rozière 1813, p. 86.
30 Du Camp 1860; Flaubert 1910.
31 Colston 1886.
32 Cailliaud 1821.
33 Belzoni 1820; d’Athanasi 1836.
34 Ababdeh is an ethnic group of the Eastern Desert. Members of the group recognize themselves as an Arab tribe, but at least some of them historically spoke a Beja language.
35 Cailliaud 1821, p. 74.
36 Cailliaud 1821, pp. 58, 60; Belzoni 1820, pp. 306‑307, 344‑345.
37 Agut‑Labordère 2018, p. 181; Agut‑Labordère, Redon 2020.
38 Chaufray 2020; Cuvigny 2020.
39 Bruce 1790, pp. 170, 181‑182.
40 On the New Kingdom networks in the Eastern Desert, see the article by I. Goncalves in this volume.
41 On the other hand, part of the resources being finite, these have of course diminished since Antiquity (see below).
42 Shadufs are water lifting devices made of a long pole hung on a frame; at one end of the pole is a counterweight, at the other end is a rope and a bucket (or a skin bag or a basket).
43 Bonneau 1993, pp. 93‑94.
44 Sāqiyat (sg. sāqia) or water-wheels are water lifting devices composed of a wheel driven by animals or humans through a mechanical system, on which buckets or pots are attached, either directly on the wheel or by means of a rope (in order to go deeper). The movement of animals or humans drives the wheel which allows water to rise up in each pot or bucket and ensures a regular and greater flow than that of a shaduf.
45 The date of the introduction of sāqia in Egypt is still debated, and based on little evidence. They were certainly known of in the Early Roman period, and maybe already in the Late Ptolemaic era. See Guimier‑Sorbets, Seif el‑Din 1997, pp. 405‑406 about the famous painting, found in Alexandria, showing the most ancient sāqia in Egypt, and probably dated to the end of the 1st century AD.
46 De Rozière 1813, pp. 85‑86; Du Camp 1860, p. 267.
47 The well of the Roman fort of Abu Zawal (anc. Rayma?) was still in use when we visited the site in January 2019, for example.
48 Tregenza 1955, p. 228.
49 Murray 1955, pp. 175‑176.
50 This highlights another factor that explains the stability of Eastern Desert networks: these were largely oriented towards the exploitation of mineral resources whose distribution has not changed over time (gold in this case).
51 Murray 1955, p. 176.
52 Klunzinger 1878, pp. 230‑233.
53 Scaife 1935, p. 65, picture 2 of plate III.
54 Scaife 1935, p. 65.
55 Tregenza 1955, p. 107.
56 Couyat 1910, p. 28.
57 Tregenza 1955, pp. 196‑197 describes it in relation to the Umm Yessar sector not far from the Qattar Roman fort.
58 De Rozière 1813, p. 95.
59 Du Camp 1860, p. 274.
60 Flaubert 1910, pp. 249‑250.
61 Klunzinger 1878, p. 229.
62 Barron, Hume 1902, p. 55.
63 Weigall 1913, pp. 70‑71.
64 See also the notes taken by M. Prickett during the survey of the hinterland of Myos Hormos made in 1978 about the spring of Ambagi, its environment, the vegetation nearby and the Beduin settlement that camped in the vicinity (Prickett 1979, pp. 270, 274, 278).
65 Pliny the Elder, HN, 6, 33, 168. The spelling “Ainos” has been challenged by several scholars, including J. Desanges in his edition in French of Pliny’s Natural history, who prefers the spelling “Tadnos”.
66 Barron, Hume 1902, p. 74; Murray 1925, plate 14, picture 2.
67 Sidebotham 1994.
68 On the correct identification of Myos Hormos with Quseir al‑Qadim, see Bülow‑Jacobsen, Cuvigny, Fournet 1994.
69 A good example of this kind of qalt is the Amusne site mentioned by Belzoni 1820, p. 336 and d’Athanasi 1836, p. 35. Unfortunately, we did not succeed in locating it precisely, but it seems to be in the vicinity of Gebel Umm Sueh. See also Ball 1912, pp. 240‑241.
70 We could not locate it precisely, but it appears to be not far from Bir Tarfawi.
71 Hobbs 2014, p. 131.
72 Bruce 1790, p. 177.
73 Couyat 1910, p. 17.
74 Couyat 1910, p. 42.
75 Barron, Hume 1902, p. 253.
76 Wilkinson 1832, p. 38.
77 Wilkinson 1832, p. 39.
78 Wilkinson 1832, p. 49. This qalt is also later described by Floyer 1887, pp. 671‑673.
79 Mac Alister 1900, p. 538 (map).
80 Mac Alister 1900, p. 537.
81 Raimondi 1923, pp. 71‑74.
82 Raimondi 1923, p. 73.
83 Tregenza 1955, p. 64.
84 The gebel (mountain) which is called Qattar by Tregenza is called Umm Disah by most of the maps and travelers’ accounts.
85 Tregenza 1955, p. 174.
86 Floyer 1887, p. 674.
87 Barron, Hume 1902, p. 253.
88 Barron, Hume 1902, p. 35.
89 Bisson de la Roque 1922, pp. 122‑123.
90 Sidebotham, Hense, Nouwens 2008, p. 308 and plate 13.1.
91 Linant de Bellefonds 1868, pp. 163‑165. His description is quite complete and mentions the cultivation of cotton, dourah and barley, plus one dum palm tree and two date palm trees.
92 Linant de Bellefonds 1868, p. 164, plate 14.
93 Sidebotham, Hense, Nouwens 2008, pp. 112‑113, picture of the site in plate 6.3. See also Sidebotham et al. 2004.
94 Belzoni 1820, p. 336; d’Athanasi 1836, p. 35.
95 Denon 1802, p. 183.
96 Denon 1802, p. 181.
97 Ball 1912, pp. 234‑236.
98 Ball 1927, pp. 114‑116; Murray 1952, p. 18.
99 Laqeita, for example, served as a fallback camp for the Mamelukes during the French Expedition in Egypt, which later necessitated its occupation by the French and their Ababdeh allies.
100 Gasse 2000, pp. 204‑205.
101 Tregenza 1955, p. 171.
102 Colston 1886, p. 501.
103 As Sidebotham, Gates‑Foster 2019, pp. 258‑259 suggests. See also Redon, Faucher 2020, pp. 40‑46 on a similar structure at Samut North.
104 Bisson de la Roque 1922, pp. 117‑121.
105 Klunzinger 1878, p. 236.
106 For example, see the picture of Bir al‑Sidd in Weigall 1913, p. 66.
107 De Rozière 1801‑1802, p. 233. Dubois‑Aymé 1802, pp. 273‑275 also gives a good description of these features.
108 Wilkinson 1832, p. 28 (map).
109 L’Hôte 1841, p. 598.
110 L’Hôte 1841, pp. 598‑599.
111 Raimondi 1923, p. 39.
112 Raimondi 1923, p. 41.
113 Raimondi 1923, p. 43.
114 Du Camp 1860, pp. 268‑269; Flaubert 1910, p. 247; Weigall 1913, pp. 65‑68; Raimondi 1923, p. 42.
115 De Rozière 1801‑1802, p. 232.
116 Belzoni 1820, p. 308.
117 Golénischeff 1890, p. 80.
118 Murray 1925, p. 145.
119 Ball 1912, p. 236.
120 Ball 1912, p. 237.
121 For example, Bruce 1790, pp. 171‑172 and Klunzinger 1878, p. 221 about Laqeita, or Belzoni 1820, pp. 303‑304 about Bir Abbad.
122 The village was excavated in January 2020 by the MAFDO; see Faucher et al. 2020.
123 O.Sam. inv. 985 (= TM 754181) published in Chaufray, Redon 2021, pp. 172‑174.
124 The site has, according to some visitors, no well and “the water lies in a cavity under the base of a hill some two or three yards north of the fort” (Scaife 1935, p. 78). This qalt was noted by the Delaware survey in 1989. A large well, likely Roman, lies 500 m to the north of the site and was probably the main water resource for the site (Maxfield, Peacock 2001, p. 236).
125 For example, Qalt al‑Aguz, where Ball mentions Ptolemaic inscriptions (Ball 1912, p. 240).
126 Cuvigny 2018, § 175.
127 Floyer 1887, p. 674.
128 Barron, Hume 1902, p. 87.
129 The concentration of salts and other soluble elements due to evaporation, as well as its mixture with organic matter and fine sediments makes it less suitable for human consumption.
130 Klemm, Klemm 2013, p. 153.
131 Bülow‑Jacobsen 2003, p. 54; Cuvigny 2003a, p. 275; Sidebotham, Gates‑Foster 2019, pp. 43‑44, 277‑278, 279‑281.
132 Ball (1927, pp. 114‑116) and Murray (1952, p. 18) were the only early scholars who understood that Laqeita and al‑Kanais were located in potential artesian zones, and took their water from the Nubian Sandstone Aquifer System.
133 Cuvigny 1997, p. 143; Cuvigny 2018, § 126, 150, 153.
134 Cuvigny 2003c, p. 381; Bülow‑Jacobsen 2003, p. 420. Bülow‑Jacobsen lists the 21 varieties grown at Persou, including lettuce, cabbage, leek, radish, chicory, but also basil, turnip and asparagus. See also Van der Veen et al. 2018, § 21.
135 The Arabic word ḥafr means “excavation” in general and does not necessarily refer to the structures discussed in this section.
136 For examples of ancient aḥfār in Sudan, see Weschenfelder 2012 and Hinkel 1991.
137 Rainfall data from Williams 2018, fig. 3.1.
138 Murray 1955, p. 180.
139 This can damage the shaft of the well and make it hard and dangerous to reopen.
140 It thus prevents water from coming inside, as shown by the wadi channel’s remains.
141 Its location, where the terrace of the Wadi al‑Ghozza and the alluvial fan of a tributary join, is optimized to capture groundwater from several watersheds.
142 This could be efficient for limiting evaporation processes.
143 Sidebotham, Gates‑Foster 2019, p. 15.
144 See infra.
145 See https://edh-www.adw.uni-heidelberg.de/edh/inschrift/HD047182 (accessed 23/11/2021). See also Bagnall, Bülow‑Jacobsen, Cuvigny 2001, p. 326 sq. for a commentary about the Siket inscription, and Cuvigny 2011, pp. 39‑41 on the slight modifications proposed by H. Cuvigny to the above-mentioned translation.
146 This reuse could partly be explained by the quality of the water supply, as the well draws its water from the Nubian Sandstone Aquifer System, which is particularly rich in water, and under pressure, allowing the water to rise naturally in the well, thus bringing it closer to the surface.
147 Cuvigny 2018, § 136.
148 The fact that several Ptolemaic sites, such as those at Persou or Compasi, have Egyptian names is probably a similar point, indicating a Ptolemaic reoccupation of previously occupied sites where the presence of water nearby (and, in the case of Bir Samut, Compasi, and Persou, gold, and hard stone) had been demonstrated.
149 See ILS 2483 = I.Portes 56 (http://inscriptions.packhum.org/text/219876?&bookid=375&location=9), which lists the Roman army units involved in the construction of cisterns in Berenike, Compasi, Apollonos Hydreuma and Myos Hormos at the very beginning of the Roman period in Egypt, and which were probably also in charge of the digging of wells in the Eastern Desert when needed (see the discussion in Cuvigny 2003a, pp. 267‑273 on the inscription).
150 See for instance the mention of “stone workers” in the inscription of al‑Kanais by Seti I (Lichtheim 1976, pp. 52‑57), or the inscription in the Paneion of the Wadi Hammamat led by a soldier and “quarrymen of the wells” (cκληρουργόc ὑδρευμάτων), in the Roman period (I.Pan 60). The same technonym is mentioned in relation to works at a hydreuma near Mons Claudianus in O.Claud. II 383.
151 O.Dios. inv. 90 is a letter from the overseer in charge of the wells which asks for reinforcements and tools to dig the well of Dios, because the rock was too hard to cut. See Bülow‑Jacobsen, Cuvigny 2007 and https://desertnetworks.hypotheses.org/702.
152 See for instance the inscription left by a certain Diodoros, sent by a king Ptolemy, to clean the well at the Paneion (i.e. at al‑Kanais): I.Paneion 12. The references in the Roman period are too numerous to be quoted here.
153 Cuvigny et al. 2010, pp. 11‑12.
154 Unpublished excavations of the MAFDO, in 2017‑2018.
155 See O.Claud. I 2, l. 4‑6, about the well of Rayma (Abu Zawal?), near Mons Claudianus and Porphyrites: indico tibi diis uolentibus aquam copiosissimam creuisse ydreuma (“I announce to you that with the will of the gods, the well has filled up, a very abundant water”).
156 For instance, water is an issue at the site of Tiberiane (Barud, near Mons Claudianus): people complain about its smell (O.Claud. IV 890, l. 16‑17: “if only the water does not smell!!”), and were obliged to leave the site due to lack of water (O.Claud. inv. 7294).
157 O.Xer. inv. 995, unpublished, mentioned in Cuvigny 2018, § 140.
158 For example, see the map “Egypt‑Eastern Desert or northern Etbai” from Floyer 1893, plate 1.
159 See for example (quoted earlier in this paper): Mac Alister 1900, p. 537; De Rozière 1813, p. 95.
160 Maxfield, Peacock 2001, pp. 42‑55.
161 Brun 2018.
162 See again the report on the excavations of the dump of Xeron Pelagos, where several layers of dredging were observed by J.‑P. Brun (Cuvigny et al. 2010, p. 11).
163 Redon 2018. The documentation of Abbad and Bir Samut makes it clear that the forts and stations of the Edfu‑Berenike road were only in use for some decades (from the 270‑260s BC to 207‑206 BC), and only intermittently, maybe in link with large hunting expeditions to the Horn of Africa.
164 More details will be provided in a paper focused on wells and in further publications about the archaeology of Ghozza.
165 They are both described in Klemm, Klemm 2013, pp. 86‑88 and 89‑92.
166 Unless the ancient miners used a sorting carried out exclusively by wind; see the discussion in Redon, Faucher 2020, pp. 335‑337.
167 Cuvigny 2018, § 175.
168 Cuvigny 2003b, p. 353. See also Cuvigny 2018, § 175‑185.
169 Cuvigny 2003b, p. 353.
170 Cuvigny 2018.
171 For the interpretation of Pliny’s sentence and the location of the stations on the road, see Cuvigny 2003b, p. 353.
172 They appear in the prescript of circulars written in Greek (study H. Cuvigny, unpublished).
173 Cuvigny 2018, § 136. The omission of the word hydreuma is common in the ostraca.
174 Cuvigny 2003b, p. 354.
175 Bagnall, Bülow‑Jacobsen, Cuvigny 2001.
176 Cuvigny 2011, p. 40 about O.Claud. I 2 and I.Did I 1.
177 Belzoni (1820) had to send men and camels to distant wells for water several times (p. 322, 325, 328, 338). He also mentions the thirst and the tensions arising from the lack of water within the expedition (p. 329‑331, 335), the death of four out of the 16 camels (p. 341, 344). Finally, the concern for water and the dangers of travel at that time are expressed in a lively passage (p. 341‑342) including a striking comparison with sea travels: “at sea, storms are met with; in the desert there cannot be a greater storm than to find a dry well”. In particular Cailliaud (1821, p. 70) mentions a conflict and a brawl that happened between Albanese and Greek workers when the expedition arrived at a well after several days without any supply.
178 See Redon 2018, fig. 1 for the Ptolemaic networks, and Brun 2018, fig. 11 and 19 for the Roman network. The maps are currently being updated by the Desert Networks project with the additions of many new sites, but this will not change the aspect of the two networks.
179 See the hypothesis of R. Ast (2018, § 29) concerning the water-supply archives found at Berenike in 2009 (O.Ber. III 274‑455), dated to the reign of Vespasian and Titus: according to him, the transport of water to Berenike under the supervision of the Roman army occurred at the same time the roads from the Valley were finally equipped (with the construction of several forts to fill in the gaps), and may be related to these activities.
180 Cuvigny 2003b, p. 333.
181 The road from Coptos to Berenike was equipped with wells around 4 BC according to De Romanis 1996, p. 224 (although Cuvigny 2003a, pp. 268‑273 discusses this point and prefers to date this opening less precisely and say that it probably took place in the immediate vicinity or before that date) and the road to Myos Hormos was already in full operation at the time of Strabo (26/25 BC).
182 Brun 2018 for a recent summary.
183 The fort of Bir Bayza, on the Coptos‑Berenike road, was quickly abandoned after its construction, and replaced by Dios/Abu Qurayyah, built 7 km to the north-west. The common hypothesis for this is that the well was not giving enough water, which shows that the reaction of the Roman army was fast and responsive (Cuvigny 2010, no. 1; Reddé 2018, § 4).
184 Cuvigny 2017.
185 As shown by the discovery of nine ovens dedicated to the mass production of bread in the fort in 2018: MAFDO excavations, unpublished.
186 Its location, in a region where the water-rich Nubian Sandstone Aquifer System can be reached, could explain it.
187 This is probably also shown by the type of amphora that caravans used in the Eastern Desert during Ptolemaic times which are, by far, the largest amphora ever attested in the Mediterranean world. See the article by J. Gates‑Foster in this volume: they could have been used to carry water.
188 Work in progress by M.‑P. Chaufray, H. Cuvigny, L. Aguer in charge of the Greek and demotic ostraca edition, and B. Redon for the network analysis. H. Cuvigny gave a first view of the Greek circulars that mention the main stations at the 26th International Congress of Papyrologists (unpublished).
189 Cuvigny 2018, § 178.
190 Apollonos hydreuma was also maybe part of the stations as early as the 3rd century BC, see above.
191 The largest one, in the Ptolemaic fortress of Bir Samut, has a capacity of 110 m3, while the tank of the Roman fort of Didymoi can contain 380 m3. See Reddé 2018, § 9.
192 It was never equipped like the Edfu‑Berenike road, and was probably mainly used under the Ptolemies by travelers, and not by official expeditions: Redon 2018, § 27. This road probably passed by Laqeita/Phoinikon which was maybe already in use in Ptolemaic times, even if it has not been proven with archaeological evidence. Thereafter, caravans had to travel for 2 or 3 days before reaching the gold-mining settlement of Daghbag/Compasi.
193 Other Ptolemaic stops with no equipment, located to the south of Berenike, in an area that was densely occupied in Ptolemaic times for its gold and also because it gave access to several water points, are also located near to a qalt (Qalt al‑Aguz) and a spring (Bir Abu Safa).
194 See above on the probably wrong identification of this structure as a ḥafr. The pottery visible on the surface is dated to the Late Roman period (Sidebotham, Gates‑Foster 2019, pp. 267‑269).
195 As described in other areas by Tregenza (1955, p. 107) and Bert (Couyat 1910, p. 28).
196 Such as those described by Bert in the north of the Eastern Desert (Couyat 1910, p. 17).
197 O.Abb. inv. 105, Cuvigny 2017, pp. 121‑122.
198 Cuvigny 2017, p. 122.
199 The ethnonym “the Arab of the desert” is mentioned in the ostracon published by Cuvigny 2020. For other mentions of Arabs in the demotic ostraca of Bir Samut, see Chaufray 2020, p. 138.
200 Unpublished material, currently under study by J. Gates‑Foster (university of Chapel Hill).
201 Chaufray forthcoming. Unfortunately, no other sources can be address here to assess the potential technological exchanges concerning water management that surely took place in the desert, between local people and Greek, then Roman settlers.
202 This aspect has been widely presented in many publications (starting with Sidebotham 1989), and some contrary opinions have also sometimes emerged. See e.g. the discussion in Cooper 2011, who acknowledges the difficulties of navigation on the Red Sea, but also exposes the problems of navigating the Nile.
203 This led to the setting up of a great military expedition from the British Indies, under the direction of General Baird, see for example the account of De Noé (1826) for more information on this matter.
204 Klunzinger 1878, pp. 235‑236 had already sensed these differences, without, of course, any mention of network analysis: “here in the desert mountains the springs or wells have become points of attraction, round which the nomadic inhabitants erect their huts until they dry up. The caravans prefer to halt in the neighborhood of the wells; the town’s people must get their water from them; they form the natural rendez vous of all the higher and lower animals that live in the desert”.
205 For more details, see Klunzinger 1878, pp. 273‑276.
206 Van Rengen 2018.
Auteurs
ERC Desert Networks; CNRS, HiSoMA (UMR 5189), Archéorient (UMR 5133), MOM, Lyon
ERC Desert Networks; CNRS, HiSoMA (UMR 5189), MOM, Lyon

Le texte seul est utilisable sous licence Creative Commons - Attribution - Pas d'Utilisation Commerciale - Pas de Modification 4.0 International - CC BY-NC-ND 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.
The Iranian Plateau during the Bronze Age
Development of urbanisation, production and trade
Jan-Waalke Meyer, Emmanuelle Vila, Marjan Mashkour et al. (dir.)
2019
Les vaisseaux du désert et des steppes
Les camélidés dans l’Antiquité (Camelus dromedarius et Camelus bactrianus)
Damien Agut-Labordère et Bérangère Redon (dir.)
2020
Chantiers et matériaux de construction
De l’Antiquité à la Révolution industrielle en Orient et en Occident
Anne Baud et Gérard Charpentier (dir.)
2020
Multidisciplinary approaches to food and foodways in the medieval Eastern Mediterranean
Sylvie Yona Waksman (dir.)
2020
On salt, copper and gold
The origins of early mining and metallurgy in the Caucasus
Catherine Marro et Thomas Stöllner (dir.)
2021
Chalcis/Qinnasrin (Syrie)
De l’âge du Bronze à l’époque mamelouke. Qinnasrin II
Marie-Odile Rousset (dir.)
2021
Kition-Bamboula VIII
Le port de guerre de Kition
Olivier Callot, Sabine Fourrier et Marguerite Yon (dir.)
2022
Networked spaces
The spatiality of networks in the Red Sea and Western Indian Ocean
Caroline Durand, Julie Marchand, Bérangère Redon et al. (dir.)
2022
Kition-Bamboula IX
Les cultes des Salines à Kition : étude des terres cuites d'époque classique
Pauline Maillard
2023
Les espaces funéraires autour d’Apamée de l’Euphrate (iiie siècle avant-viiie siècle après J.‑C.)
Catherine Abadie-Reynal (dir.)
2023
Les théâtres antiques et leurs entrées
Parodos et aditus
Djamila Fellague et Jean-Charles Moretti (dir.)
2024