538 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1963 



blocks are carried away for use in modern constiniction, debris soil 

 is being removed by farmers for their fields, and in some instances, 

 the site has been leveled, plowed, and incorporated into a modern field 

 system. It is worth noting in passing that these forces of destruction 

 have made it impossible in many parts of the sites to determine the 

 stratigrapliic relationship between buildings, since the intervening 

 layers of debris have been removed. 



Because of erosion and other agents of destruction, many details 

 of the ancient sites can be seen on the surface. Most of the buildings 

 were constructed with foundations and lower courses of stone, and a 

 variety of masonry types and dressing techniques appear in each site 

 and occasionally in the same building, suggesting different periods of 

 construction. The superstructure is usually represented by a pile of 

 disintegrated mud brick, but often mud-brick walls and even in- 

 dividual bricks can be distinguished (pi. 7:1). On the larger and 

 earlier sites, the surface is strewn with literally hundreds of thousands 

 of sherds which have eroded out of the debris. Architectural frag- 

 ments, incense burners, an occasional fragment of sculpture, inscribed 

 pieces, and obsidian points and sickle blades are also found. 



Perhaps the most surprising discovery of the survey is the lack of 

 townsites in the main wadi. Today the main wadi contains perhaps 

 90 percent of the entire population of the Hadhramaut drainage. 

 Because of its present large population, it is reasonable to expect a 

 proportionately heavy population in antiquity and to find a heavy 

 concentration of sites in the main wadi. Yet only one townsite and a 

 trace of a second were found; the former is located on the gravel- 

 strewn talus, while the skimpy remains of the latter are situated on 

 the silt floor of the valley. How can this apparent lack of ancient 

 settlements be explained ? Wliile the collected data have not yet been 

 thoroughly evaluated, it seems likely that there may have been many 

 sites in the main wadi in antiquity which were destroyed within the 

 last thousand years or so by (1) the construction of irrigation installa- 

 tions, and (2) the erosion of the valley floor. With regard to the 

 former, it should be noted that large sections of the main wadi have 

 been cultivated using seil irrigation techniques. This method of iiTi- 

 gation, which has been used throughout the Islamic period, requires 

 the leveling of fields, the cutting of canals in the silt floor, and the erec- 

 tion of earthen dykes around each field plot. Thus everything, in- 

 cluding ancient building remains, was leveled. From many of the 

 dykes still in use and from others long abandoned, pre-Islamic pottery 

 is eroding out, while in the immediate vicinity, there is not a trace of 

 a pre-Islamic structure. It was from such a dyke that the only trace of 

 pre-Islamic pottery on the silt of the main wadi was found. It should 

 also be noted that the silt beneath this dyke now stands 2 meters above 



