An unpublished shelter with prehistoric engravings of a possible late Pleistocene date in the North-central Sinai (Egypt) Andras ZBORAY* In the year 2000 the Egyptian naturalist Gabriel Mikhail was led by local Bedouin to a large shelter containing a number of petroglyphs. The few taken photos showed some remarkable animals engraved in what appeared to be raised relief, in a style totally different from the known petroglyphs of the Eastern and Western Deserts of Egypt. In January 2001 the author was privileged to have been offered the opportunity to visit the site, and to take more detailed photographs of the panels. At the time it was assumed that the site must have been recorded previously. However, a review of the rather sketchy literature of the region seems to indicate that the shelter was never published, it remained unknown even to specialists dealing with the prehistory of the Sinai Peninsula (Emmanuel Anati, pers. comm.). This is most probably due to its location, about 140 kilometres to the East of the Suez Canal and the town of Ismailia, and about 80 kilometres south of the Mediterranean coast (Fig. 1). The site itself is located about five kilometres upFig. 1. Location map. Fig. 2. Shelter overview. stream in a wadi (dry watercourse) draining the central limestone plateau of the Sinai, which abruptly rises from the flat terrain to the north and west. This area was the scene of military activity practically throughout the second part of the Twentieth Century, with the post 1973 demarcation line running very close to the site location. Even presently it is a restricted military area, as it must have been for the past 50-60 years, with only local Bedouins having unhindered access to the valley hiding the site. The shelter is located about 300 metres upstream in a lateral wadi of the main valley, hollowed out under a dry waterfall. There is a dense scatter of flint debitage and a few occasional finished artefacts on the surface of the gravel terraces flanking all sides of the lateral wadi as well as the main valley surrounding the confluence. The noted artefacts were all based on the blade technology, some displaying fine pressureflaking retouch, a feature common to advanced blade industries from the late Palaeolithic well into the Neolithic. On the far side of the main valley, on a local high point of the gravel terrace, there are several circular stone structures, however their association with the artefact scatter is doubtful. While admittedly the visit was briefand no thorough survey was made of the area, surface ceramics were conspicuous in their absence. The shelter itself is approximately 5-6 metres high, with a similar depth at the centre, and a width of 12-15 metres, located about halfway up the 20-metrehigh dry waterfall, accessible by climbing a scree slope that has formed in front of it (Fig. 2). The floor plan is approximately that of a semicircle, with an arching roof that merges into the vertical rear wall. Engravings are found in widely spaced groups along the rear wall, from a height of approximately 50 centimetres above the floor to above human height. As one enters the shelter after ascending the scree slope on the right, two conspicuous engraved animals are immediately visible. Both are executed with deeply grooved outlines, the outer edges of which were further reduced and the edges of the animals rounded, * Deneuer u. 48/D 1121 Budapest (Hungary) andras@[jexpeditions.com SAHARA 2312012 163 producing a raised relief (Fig. 3). Either through repeated touching or already at the time of their making, the bodies of the animals were smoothed, further enhancing the relief effect. While some of the rounding and smoothing may be the result of weathering, the pattern exists throughout the site, leaving no doubt as to its intentional origin. The two animals face left towards the interior ofthe shelter, the rounded muzzle of the one on the left, and the rounded belly of the one on the right suggest an equine species. Further left there are several more quadrupeds, executed in the same technique. One is unquestionably an ibex, while the other more weathered ones may be tentatively identified as either ibex or gazelle (Fig. 4 and 5) The most striking scene in the shelter is located in the central part, close to eye leveL It consists of several quadrupeds, of which only one (also apparently an equine) is readily recognisable, the others have mostly weathered away. To the left of the animals there is the torso of a female figure depicted in profile, with very prominent 164 SAHARA 2312012 Fig. 3. Equines in raised relief. Fig. 4. An ibex. Fig. 5. Group of quadrupeds in relief. Fig. 6. Female torso with quadruped. Fig. 7. Female torso close-up. breasts and buttocks, lacking both the head and lower limbs. This is clearly not the product of weathering or other damage, the delimiting grooves indicate that it is a complete and intentional composition (Fig. 6 and 7). The torso is about 10 centimetres high, while the animal to the right about 20 centimetres wide, a good approximation for the size of all the other animal depictions in the shelter. Left of the torso there are further quadrupeds, more weathered than the ones described above, but executed in the same style and technique. One is an ibex, the others have body proportions suggesting gazelle or antelopes (Fig. 8, 9). The right of the shelter is covered by a large complex panel ofnumerous figures, starting from below eye level and reaching about a metre higher. These depictions differ from the previously discussed ones. Their outlines are made with much shallower grooves, and there is no rounding of the edges or smoothing of the interior of the bodies. However some of the animals depicted look exactly like the ones executed in raised relief, including a number of equines (Fig. 10 and 11). Near the centre of the right panel, it is possible to recognise a second female torso, cruder but otherwise similar to the first one except in the execution technique. It is surrounded by various animals, including an ostrich imme- diately above (the only clear example in the shelter), a readily identifiable oryx, and more surprisingly a dromedary (Fig. 12). Near the top of the panel, there is a row of weathered quadrupeds, which on close scrutiny are revealed to be dromedaries, accompanied by another oryx (Fig. 13). The appearance of the petroglyphs in the shelter, both in style and weathering, suggests an age of great antiquity, apparently at odds with a historic date indicated by the camel depictions. However the presence of Camelus species in the Egypt-Sinai-Negev region is demonstrated from the late Pleistocene, though not the domesticated dromedary but Camelus thomasi (Veerle Linseele, Katholieke Universiteit Leuven, pers. comm.; see also Gautier, 1966; Grigson, 1983; Uerpmann, 1987). Similarly, the depicted equines may be identified either as Mrican wild ass, Equus africanus, or the Asian wild ass, Equus hemionus, the historic ranges of which could have overlapped in the Sinai (Uerpmann, 1987) in the late Pleistocene and early Holocene. The remainder of the depicted fauna, ibex, ostrich, oryx, gazelle and antelopes are either endemic species, or their historic range encompassed that of the Sinai Peninsula since at least the late Pleistocene. The style of the depictions bears no similarity to other published sites from the Sinai (Anati, 1979), and is also very different from any depictions known to the author from the vicinity ofthe Nile valley or the deep deserts of Egypt. It is necessary to look further East to find any analogies. The site of Kilwa, located in present day Saudi Arabia, just across the border from Jordan, is approximately 400 kilometres to the East of the described shelter, not substantially farther than the Nile valley some 300 kilometres to the West. Already noted by Gertrude Bell in 1914 (Hors- Fig. 8. Group of quadrupeds. Fig. 9. Ibex. Fig. 10. Equines (?) in outline, lacking relief. Fig. 11. Group of equines. field, 1943), the site of Kilwa was first explored systematically by Leo Frobenius and Hans Rhotert in 1934 during DIAFE XII (12 th Deutsche Inner-Afrikanischen Forschungs-Expedition). The locality contains petroglyphs and archaeological remains from several periods. On one particular hill, named "Horsfeldberg" by Rhotert, there are a group of engravings which based on patination, superimpositions and style were considered to be the oldest in the area, and assigned to the Mesolithic (Rhotert, 1938). It was noted that no other rock art locality of the area contained this type SAHARA 2312012 165 Fig. 12. Group of animals and female torso in outline only. Fig. 13. Dromedaries and an oryx. of engravings. Of particular interest are some ibex, which bear great similarity to the ibex depicted in the described Sinai shelter. More importantly, Rhotert noted a small dromedary close to a "Mesolithic" ibex both having similar patination' suggesting contemporarity (Rhotert, 1938: PI. 15.2). Another larger dromedary (Rhotert, 1938: 424) is in the same style, and was also assigned to the same period. Anati (1979) established a relative chronology for the rock art ofthe Sinai-Negev region, and included the early Kilwa depictions in his ｓｴｹｬｾ I, Early Hunters. Based on supenmpositions, Style I was ､･ｭｯｮｳｾ｡ｴ･､ to be the oldest in the established sequence, and was tentatively dated to "before the beginning of the 6th millennium B.C." Even on first impression, the similarity to the Kilwa Ｎｴｨｾ aspetroglyphs would ｳｾ･ ｴ signment of the descnbed Smro. engravings to the Early Hunters style, however a further confirmation comes from South-<:entral Arabia. In the monumental four volume work detailing the rock art finds of the 1951-52 Philby-Rickmans-Lippens expedition to Central Arabia, there is one panel attributed to the style of the Early Hunters which contains a single headless and legless torso of a female figure that seems to be a very good parallel for the female figures in the Sinai cave (Anati, 1974: fig. 174). In summary, the evidence suggests that the described Sinai en- 166 SAHARA 2312012 gravings were made by a hunting society that seems to have roamed across the southern Levant and the Arabian Peninsula at or before the first half of the 6th millennium B.C., of which at present this shelter is the westemmost known manifestation. The depicted fauna would permit any time period back to the late Pleistocene. The absence of any surface ceramics in the vicinity of the shelter otherwise rich in flint artefacts with a pressure flaking retouch, gives further support to a preNeolithic terminus ante quem for the engravings. While ｣ｯｮｶ･ ｾｯｮ｡ｬ chronologies would not pemut the existence of any pictorial art in the region prior to the early Holocene, the firm dating of a panel of engravings at Qurta in the upper Nile Valley to the late Pleistocene (minimum age 14-16 thousand years Bp, ｈｾＺｲｧ･ et al., 2011) raises the tantalizmg possibility that some ?f the. knoW? rock art of the region, mcluding this Sinai shelter, may belong to periods older than previously thought. Further support to this hypothesis is provided by the finding ofheadless human figures (likely female) with pronounced buttocks, but lacking any other anatomical features, at the dated Qurta locality (Huyge et al., 2007). Such "femmes sans tete" depictions are also an integral part of the European late Palaeolithic rock art repertoire (Bosinski, 2011). References ANATI E., 1974. Expedition Philby·Ryck· mans-Lippens en Arabie. Premiere partie: Geographie et arcMologie. 1bme 3: Rock·Art in Central Arabia. Vol. 4: Corpus of the Rock·Engravings. Parts III & N. ｌ ｯ ｵ ｶ ｾ ﾭ la-Neuve. Publications de l'Institut Orientaliste de Louvain, 6. ANATI E., 1979. L'Arte Rupestre del Negev e del Sinai. Milano: Jaca Book. BOSINSKI G., 2011. Femmes sans rete. Une ic6ne culturelle dans ['Europe de la fin de l'ere glaciaire. Paris: Errance. GRIGSON C., 1983. A very large camel from the Upper Pleistocene of the Negev Desert. Journal ofArchaeological Science, 10: 311-316. GAUTIER A., 1966. Camelus thomasi from the Northern Sudan and its Bearing on the Relationship C. thomasi-C. bactrianus. Journal of Paleontology, 40: 1368-1372. HORSFIELDA., 1943. Journey to Kilwa, Transjordan. The Geographical Journal, Vol. cn No. 2 (August 1943), p. 71-77. HUYGE D., M. AUBERT, H. BARNARD, W. CLAES, J.C. DARNELL, M. DE DAPPER, E. FIGARI, S. IKRAM,A. LEBRUN-NELlS AND I. 1'HERASSE, 2007. Lascaux along the Nile: Late Pleistocene rock art in Egypt. Antiquity, Vol. 81 No. 303 (September 2007), Project Gallery [http://www.antiquity. ac.uklprojgalllhuyge3131] HUYGE D., D.AG. VANDENBERGHE, M. DE DAPPER, F. MEES, W. CLAES AND J.C. DARNELL, 2011. First evidence of Pleistocene rock art in North Africa: securing the age of the Qurta petroglyphs (Egypt) through OSL dating. Antiquity, Vol. 85, No. 330 (December): 1184-1193. RHOTERT H., 1938. Transjordanien. Stuttgart: Strecker und Schroder. UERPMANN H.·P., 1987. The ancient distribution of ungulate mammals in the Middle East. Wiesbaden: Dr. Ludwig Reichert.