a region/city in the proximity of the Rub al-Khali sand desert of southern Arabia or a location to the south of it somewhere near eastern Yemen or western Oman (18°15′31″ N, 53°39′05″ E). Medieval Arabic sources point to Ubar as a land in eastern Yemen. Early Islamic Arab historians suggest that the region was connected to the land of the Mahra, near al-Shihr, and al-Akhaf (Tkach 1897, pp. 1832–1837; 1913a, pp. 1073–1074). Originally, the region reportedly was well irrigated, forested, and had many fortresses and palaces. In the local tradition, destruction was brought down on the people by their wickedness, and the ruins became choked with sand from great winds or drought. Eventually, the area was abandoned. Several authors tie the Ubar area to both the Mahra and the long-lost people of Ad. The assumption that Ubar was a single city comes from the more fanciful embroideries of the Thousand and One Arabian Nights, particularly the stories “The Keys of Destiny” and the “City of Brass.”

Other than St. John Philby in the early 1930s (1933, pp. 157–180), few Westerners have searched for Ubar/Wabar. A key to the mystery is in the term Iobaritae used by Ptolemy (6.7.24) to define a tribal group living inland of the region known as Sachalitai (Stevenson, 1932, pp. 137–140). Both terms appear in modern Dhofar on copies of Ptolemy's map and in the Periplus of the Erythraean Sea, among other terminology consistent with a locality in the northern Indian Ocean (Casson, 1989, pp. 170–173). Central to the matter is identifying the Iobaritae with the Islamic term Ubar. J. Tkach thought the people of Ubar lived in deep antiquity and had disappeared prior to the coming of Islam (1913a, p. 1074). In the nineteenth century, Western scholars argued that the Iobaritae were the later Ubars (Tkach, 1897, p. 1836). Scholars are divided today on the connection.

With the first investigations at Dhofar in the late 1950s, the site of Khor Rori (long identified with Moscha of the Periplus account), was confirmed to be ancient SMHRM (Albright, 1982, cf. the arguments on pp. 3–10; Doe, 1983, pp. 147–150). Its gateway inscription in Epigraphic South Arabic specifically mentions the S'KLHN frankincense district (Albert Jamme in Albright 1982, pp. 42–45; Pirenne, 1975; Beeston, 1976) and supports the term Sachalitai used by both the Periplus and Ptolemy. Based on the archaeological survey on the Salalah plain (1992–1995), it is debatable, however, if the term Moscha Limen used by the Periplus is to be identified with the Khor Rori location (see Doe, 1983, p. 21) because SMHRM = Moscha is largely conjectural. Ptolemy's harbors and settlements on the Salalah plain can now be more readily identified, based on the current archaeological survey of this region carried out by Juris Zarins. At least seven major sites from Khor Rori to Raysut (and perhaps Mughsayl) must be considered in reference to Ptolemy's locations.


Ubar, a region/city in southern Arabia. Some debate remains over the exact location and the ethnicity of its inhabitants.

view larger image

It appears that Ptolemy's placement of the Iobaritai in the hinterlands behind the Sakalan region is also accurate (contra Groom 1994, pp. 207–208). The three localities (Iula, Marimatha, Thabane) identified with the Iobaritae may be tied into the Mahra occupation of the region. These identifications are contingent on solving problems of scale and on the distortion inherent in all copies of Ptolemy. (For the contemporary distribution of the Mahra, see Dostal, 1967, and Lonnet, 1985; as a contemporary term in the classical period, see Müller, 1991.)

Bertram Thomas's observation that the road to Ubar (1932, p. 161; see also Groom, 1994, p. 208) headed into the Rub al-Khali was the basis for the recent investigations by Zarins and the Jet Propulsion Laboratory Scientists Ron Blom and Robert Crippen into the location of Ubar. LANDSAT/SPOT/THEMATIC MAPPER remote-sensing images supplied by the Jet Propulsion Laboratory of the area Thomas noted recorded a long route beginning near Wadi Mitan through the southern Rub al-Khali. Projecting the route northward, the destination would have been either Ayun/Layla in the central Tuwaiq Mountains or the oasis of Jabrin on the northern edge of the Rub al-Khali (Philby, 1933, pp. 86–106). Archaeological work in the Rub al-Khali discredited the theory that large, permanent settlements belonging to the Ubar region would have been located in the sands (Edens, 1988). Tracing the tracks Thomas discovered back to the nearest ancient permanent source of water led to the discovery of Shisur (Ar., “cleft,” or “fissure”). Known to explorers since Thomas (1932, p. 136) and Wilfred Thesiger (1984, p. 100), and even visited by archeologists in 1973 (Pullar, 1974, p. 41; 1975, pp. 49, 71–73) the site and region were not systematically investigated until 1990–1992.

A brief surface examination of the Shisur promontory led to a detailed study of the site and region by Zarins in 1991–1992 as director of the Transarabian Expedition. Data from this study and the excavations (1992–1994) now confirm that the site is most likely Marimatha, the largest settlement on Ptolemy's map, in the territory of the Iobaritae. The Shisur site, apparently a collapsed limestone cavern, is now a large sinkhole. Earthquake activity ruptured the aquifer beds (Umm ar-Radhuma A–B) and water flows to the surface in several springs. Fracturing and sink-hole collapse have been seen at Shisur and in the surrounding area. The events shaping the region's current geomorphology are dated to the last phases of the Pleistocene. Wadi Ghadun, 2 km (1 mi.) west of Shisur, represents the region's other major geomorphological feature. The now dry stream was a major feature during the Pleistocene and Early Holocene, flowing past Shisur and skirting the Rub al-Khali, 25 km (16 mi.) to the north.

A preliminary assessment of the site's history is based on the excavation of 3-meter squares laid in a large grid. The earliest human activity belongs to the Neolithic period (phase III, c. 3500–2500 BCE), although more than twenty-five sites of the period are known around Shisur that extend occupation back to about 6000 BCE (phase I). Possible Upper Paleolithic (100,000–40,000 bp) materials may be located to the east. Nondescript stone tools at Shisur may belong to the Bronze Age (2500–1000 BCE), but any formal buildings date to the Iron Age (c. 1000–0 CE). Ceramics only appeared in the region, at Dhofar (Tkach, 1913b), in the Iron Age—in contrast to Yemen (2500 BCE) and northern Oman (4500 BCE), so that investigations are focusing on the structural and material culture sequences of the Iron Age and later. [See Yemen; Oman.]

The earliest fortress at Shisur may be Early Iron Age. A small enclosure protecting the springs was built with semi-dressed stone walls with interior partitions. Both detailed stratification and the material culture suggest that the fortress was enlarged in the Iron Age and that occupation peaked in the early first centuries CE. The walls of the fortress were then enlarged, more interior rooms were added, and a proper gate and additional towers were built. The material culture resembles that of both Yemen (imported raw chlorite schist, sandstone, and basalt, which were crafted into finished products) and northern Oman (Parthian red wares connected to Sohar). [See Sohar.] The distinctive red-burnished wares with dot-and-circle/rouletting motifs have yet to be identified outside Dhofar. Other Hellenistic wares with appliqué point to connections with eastern Saudi Arabia, specifically the Jabrin, Thaj, and Dammam regions. [See Thaj.] After a four-century hiatus (c. 400–800 CE), probably caused by earthquake activity that destroyed a large portion of the site, or economic collapse the complex was formally reoccupied in the ῾Abbasid period and finally abandoned by 1400. Excavation at the bottom of the sinkhole revealed more than 5 m of occupation, all attributable to bedouin use since 1400.

The region surrounding Shisur parallels the focus of the central formal complex. Especially to the east, campsites and small-scale satellite occupation can be seen for more than 10 km (6 mi.) along an arc of small hills. At Hailat Araka, 40 km (25 mi.) away, a similar occupation pattern may be present: limited examination yielded principally large-scale Neolithic and possible Bronze Age occupations. Irrigation channels and fields reported at Shisur are now obscured by modern oasis and farm activity (Thomas, 1932, p. 131). This situation may parallel the ῾Ain Humran complex on the Salalah plain, where irrigation channels and field walls can still be seen.

As a whole, the Shisur region, because of its perennial water supply, was the focal point of commerce and agriculture. Based on Ptolemy's observations, it appears that the Iobaritae were the ancestors of the Mahra peoples inhabiting this plain and conducting incense trade across the Rub al-Khali and elsewhere. The smaller satellite sites with perennial water dot the backslope of the Nejd. Ptolemy's sites of Iula and Thabane may be Andhur and Mudai (or Habarut), respectively, in Dhofar. Following the collapse of the incense trade and possible topographical changes in the region caused by climate and/or earthquake, the Mahra retreated to the mountains to the south sometime after the tenth century (Dostal, 1967, pp. 123–135, esp. fig. 19). The medieval Arab trade network through the region was destroyed by the Portugese in the fifteenth century (Guest, 1935, pp. 407–408).


  • Adams, Robert McC., et al. “Saudi Arabian Archaeological Reconnaissance 1976.” Atal 1 (1977): 21–40. For the route beginning near Wadi Mitan through the southern Rub al-Khali, see pp. 28–30 and pl. 3.
  • Albright, Frank P. The American Archaeological Expedition in Dhofar, Oman, 1952–1953. Publications of the American Foundation for the Study of Man, vol. 6. Washington, D.C., 1982.
  • Beeston, A. F. L. “The Settlement at Khor Rori.” Journal of Oman Studies 2 (1976): 39–42.
  • Bibby, Geoffrey. Preliminary Survey in East Arabia, 1968. Copenhagen, 1973. For the archaeology of the route beginning near Wadi Mitan through the southern Rub al-Khali see pp. 48–59.
  • Casson, Lionel, ed. and trans. The “Periplus Maris Erythraei”: Text with Introduction, Translation, and Commentary. Princeton, 1989.
  • Doe, Brian. Monuments of South Arabia. New York, 1983.
  • Dostal, Walter. Die Beduinen in Südarabien. Vienna, 1967.
    For the contemporary distribution of the Mahra in the classical period. See also Lonnet (below)
  • Edens, Christopher. “The Rub al-Khali ‘Neolithic’ Revisited: The View from Nadqan.” In Araby the Blest: Studies in Arabian Archaeology, edited by Daniel T. Potts, pp. 15–43. Carsten Niebuhr Institute Publication, 7. Copenhagen, 1988.
  • Groom, Nigel. “Oman and the Emirates in Ptolemy's Map.” Arabian Archaeology and Epigraphy 5 (1994): 198–214.
  • Guest, A. R. “Zufār in the Middle Ages.” Islamic Culture 9 (1935): 402–410. Discusses the medieval Islamic evidence for the route beginning near Wadi Mitan through the southern Rub al-Khali, p. 408.
  • Lonnet, Antoine. “The Modern South Arabian Languages in the P.D.R. of Yemen.” Proceedings of the Seminar for Arabian Studies 15 (1985): 49–55.
    For the contemporary distribution of the Mahra in the classical period. See also Lonnet (below)
  • Müller, W. W. “Nahra.” In The Encyclopedia of Islam, new ed., vol. 6 pp. 80–84. Leiden, 1991.
    See this for Mehra as a contemporary term in the classical period
  • Philby, H. St. John. The Empty Quarter. London, 1933.
  • Piesinger, Constance M. “Legacy of Dilmun: The Roots of Ancient Maritime Trade in Eastern Coastal Arabia in the Fourth/Third Millennium B.C.” Ph.D. diss., University of Wisconsin, Madison, 1983. For the route from Wadi Mitan through the southern Rub al-Khali, see pp. 191–206.
  • Pirenne, Jacqueline. “The Incense Port of Moscha (Khor Rori) in Dhofar.” Journal of Oman Studies 1 (1975): 81–96.
  • Pullar, Judith. “Harvard Archaeological Survey in Oman, 1973, I: Flint Sites in Oman.” Proceedings of the Seminar for Arabian Studies 4 (1974): 33–48. For Shizur, see p. 41.
  • Pullar, Judith. “A Selection of Aceramic Sites in the Sultanate of Oman.” Journal of Oman Studies 1 (1975): 48–88. For Shisur, see pp. 49, 71–73.
  • Stevenson, Edward Luther. Claudius Ptolemy: The Geography. New York, 1932.
  • Thesiger, Wilfred. Arabian Sands (1959). London, 1984. For the discovery of Shisur, see p. 100.
  • Thomas, Bertram. Arabia Felix. New York, 1932. For the discovery of Shisur, see p. 136.
  • Tkach, J. “Iobaritai.” In Paulys Realencyclopädie der classischen Altertumswissenschaft. Stuttgart, 1897.
  • Tkach, J. “Wabar.” In The Encyclopaedia of Islam, pp. 1073–1074. Leiden, 1913a–.
  • Tkach, J. “Zafār.” In The Encyclopaedia of Islam, pp. 1187–1190. Leiden, 1913b–.
  • Zarins, Juris, et al. “Saudi Arabian Archaeological Reconnaissance 1978” Atlal 3 (1979): 9–42. For classical-period settlements on the route beginning near Wadi Mitan through the southern Rub al-Khali, see pp. 26–29, 31–35 and pls. 14–18.

Juris Zarins