Because they were threatened by agricultural development, we had been encouraged to examine these sites by Dr. M. Bafaqih, the Director-General of the General Organisation for Antiquities, Museums and Manuscripts (GOAMM) of Yemen and in 1993, the present writer and M.C.A. Macdonald had made an initial visit to the same area. Our fieldwork in 1994 was intended to expand on these initial observations and eventually to lead to excavations.

The coastal area in the neighbourhood of Aden and Abyan forms a sandy fertile plain, with several wadi systems which drain from the highlands to the Gulf of Aden and the Arabian sea. The area is similar to some parts of the Tihama along the Red Sea coast and some Arabic geographers regard the southern coast as part of the Tihama.

The Greek, Latin and Arabic sources give a certain amount of information about the Aden hinterland and the Abyan district. Of early sources Ptolemy is especially important, as his map (ca AD 90-168) marks several places to the east and NE of Aden including Bana, which has been identified with Wadi Bana, one of the principal channels flowing through Abyan.

AI-Hamdani, writing in the 10th C. AD, is the most informative of early Islamic geographers on the Aden area and Abyan, devoting a considerable amount of attention to its wadis, tribes and settlements. Among the main tribes of the area he mentions Dhu Asabih on numerous occasions: they were important in the region in the pre-Islamic period and they were found in Abyan and Lahej in the 10th century. They remained a significant tribe in the country north of Aden down to the present century when R.B. Serjeant studied them: in the tenth century, Bani Majid were also found in Aden and Lahej. According to al-Hamdani, Khanfar on the Wadi Bana was the main town of Abyan, where both the Asabih and the Bani Majid lived. Khanfar repeatedly emerges in the Arabic sources as the principal town of Abyan: thus Yaqut (ca 1224) also refers to Khanfar as a madina of Abyan, and implicitly more significant than other places he names. He also preserves place names in Abyan not given elsewhere, speaking of specific husun and qala’at. Abu Muhammed al-Taiyib b. Abd Allah Makhrama in 928/1521-2 also mentions Khanfar as one of the famous towns of Abyan along with al-Mahal, which we are inclined to associate with the large site of al-Qarn, near Zingibar. However, of the beautifully built large mosque at Khanfar to which Makhrama refers, nothing remains. Indeed, it seems that other buildings of some interest have been lost in the region as Husayn b. Salama (d. 402/1011), a wazir for the Ziyadids, is said to have built mosques from Hadhramawt to Aden, Abyan and Lahej: none of these appear to survive in the Aden and Abyan areas.

The first significant archaeological fieldwork in the Aden hinterland at Kawd am-Sayla, al-Habil (NW of Lahej) and at Zinjibar was initiated in 1941 and published by A.Lane and R.B.Serjeant. They recognised the evidence of glass production and also the importance of Far Eastern trade during the Islamic period, marked by the presence of Chinese pottery. The first extensive archaeological survey to include sites around Aden and Abyan was undertaken by G.L. Harding in 1959-60: however, the security situation was bad and many areas were closed, so Harding’s freedom to move was restricted in the Western Protectorate. Nevertheless, his report along with that of Lane and Serjeant remains the foundation of archaeological fieldwork in the area. Subsequently, D.B.Doe published further observations on the Aden and Abyan areas, providing a catalogue of sites to east and west of Wadi Bana. This was the sum of published research on the archaeology of this area when we began work in 1994, although there has been a growing interest in the district among archaeologists working in Yemen in the past two years.

As a general observation, the Islamic sites north of Aden and in Abyan are remarkable for their size and number, and several show signs of industrial activity, probably connected with glass production. The impression conveyed by the size of these tells is one of sustained and significant settlement in the mediaeval period in coastal Yemen. Furthermore, the evidence of eastern trade marked by the presence of Chinese celadon and blue and white porcelain, which Lane and Serjeant first commented on, is widespread at the sites we examined. This observation fits into a much wider pattern of trade linking the Arabian coasts with the Far East, which is becoming better understood as a result of excavations elsewhere in the peninsula, especially at Julfar in Ras al-Khaimah. It is also clear from surface remains that Mocha on the Red Sea coast of Yemen will prove to be a significant site in terms of eastern trade.

Our re-examination of Kawd am-Sayla near Shaikh Uthman confirmed the importance of the site among those in the vicinity of Aden. It is a large tell covering an area of some 350 m x 250 m and it is about 7 m high, with ceramics suggesting a l3th-l6th century date: our dating supports that reached by D. Whitcomb after a review of the ceramics from the area. Kawd am-Sayla has plentiful evidence of glass manufacture with glass slag scattered over the site and ashy deposits from industrial activity. Serjeant suggested that Kawd am-Sayla was identical with al-Lakhaba, a place mentioned several times by Makhrama as a populous village with shops and presses for oil or sugar and from where glass and bricks were transported to Aden.

Nearby is Kadumat am-Shaibi, a still larger site on the east side of Wadi alKabir, with the main area of tells extending for about 350 m although lower mounds and sherd scatters continue for a much greater distance, and the total occupation area extends for as much as a kilometre along the Wadi. There were large quantities of Islamic pottery on the surface, once again of ca 13th-16th century date, and also glass scattered over the site and on the higher southern tell there were reddish baked brick walls.

At Subr, on the road from Shaykh Uthman to Lahej, we did little as its extensive tells had been examined by Harding and Doe, who had recognized its importance as a pre-Islamic site. Until the summer of 1992, the tells on the eastern side of the road were undamaged, whereas those to the west had been built upon many years before. By 1993, one tell had been partly cut away, exposing a section with a mass of unglazed pottery and some ash, but showing no evidence of floors or structures. The site deserves attention as a matter of priority since development is threatening its integrity.

In Abyan, the tells were similarly large and of approximately the same date as those north of Aden. A major site apparently not reported hitherto was found at Shaikh Salim, a village to the east of Zinjibar in Abyan. The Shaikh Salim tell was free of buildings and stood 10 m or more high, with baked brick wall traces and plentiful evidence of industrial activity, indicated by scatters of kiln bricks and glass, and Islamic and Far Eastern imported ceramics of ca 13th-16th century date.

Al-Qaru is to the north of Zinjibar, another very large tell about 1Gm high, and spreading for some 600 m. Once again there are numerous unglazed and glazed Islamic sherds, Far Eastern imports and glass frit.

Further inland we examined al-Tanya, which is referred to by Yaqut in the 13th century. The modern village of al-Tanya stands on a hill overlooking fields where, some fifteen to twenty years ago, the GOAMM excavated a hoard of gold coins. To the north of the village was a qubba dedicated to a local wali, Shaikh Ahmad b. Jafri, built in the conventional regional style for funerary architecture, with a steeply curving conical dome. Beside this qubba was a low area of archaeological deposit measuring about 150 m x 80 m, with numerous Islamic sherds, apparently of the same general date as the ceramics at Shaikh Salim and other sites we examined.

Just as in mediaeval times, Khanfar is a major town in Abyan and is rapidly expanding with a great deal of new construction. This has caused immense damage to the underlying archaeological remains. This expansion has transformed the place since Harding photographed it in 1959-60.

The foundations sunk for the modern buildings had the accidental advantage of exposing the stratigraphy below the present ground surface down to at least 4 m depth. There was ashy deposit containing glazed and unglazed Islamic pottery. The ceramics suggested a somewhat earlier date for Khanfar than those at other sites which fits with al-Hamdani’s reference to the town in the tenth century.

Our survey underlined the importance of the archaeology of the region both in terms of the Islamic period in the Yemen and more broadly, in terms of the Arabian littoral, the study of which has been very patchy and sporadic so far. Furthermore, it seems quite clear that the future study of the sites in Aden and Abyan will have wider ramifications for understanding of Indian Ocean and Far Eastern trade connections with Arabia.

November 1994