ARCHEOLOGY IN SOUTH ARABIA — ^VAN BEEK, COLE, AND JAMME 529 



weathered material had been chipped away and the flint nodules 

 recovered, but as fresh limestone was reached, the flint miners had 

 to content themselves with flaking what they could from the embedded 

 nodules. Trimming debris could be found for a considerable distance 

 down the slope below the site of mining operations. 



Flint or chert was similarly trimmed from detached blocks of lime- 

 stone. In some cases where the block was sufficiently small, the matrix 

 was apparently flaked away to bare the desired nodule. In any event, 

 it is not uncommon to find large trimming flakes that are part chert 

 and part limestone. 



Limestone hencfi sites (fig. 2-D). — Stone implements, flakes, and 

 debris commonly occur on the bench at tlie top of the lower cliff- 

 forming limestone. In places, especially at wadi junctions, this bench 

 becomes quite wide forming a large, very flat area. These features, 

 which provide a flat surface free of limestone rubble, a good view 

 into both wadies, and easy access to the wadi floor, seem to have been 

 a favorite camping situation for certain of the prehistoric inhabitants 

 of the Hadhramaut. They usually contain artifacts in sufficient 

 abundance to warrant collection. Occasionally broad, flat, bench 

 localities were found which had few or no artifacts, but generally 

 this seems to have been the result of a local deficiency of raw material 

 for toolmaking. 



The artifacts found on the bench sites appeared — superficially at 

 least — to be similar to those collected from the plateau, although 

 occasionally material reminiscent of that found on low-lying features 

 was encountered. The artifacts are usually darkly patinated. It was 

 noted, however, that more trimming debris is to be found on these 

 sites than on the plateau, presumably reflecting easy accessibility of 

 raw material from the talus which is always close by. 



Although the bench sometimes narrows to a mere ledge or disappears 

 altogether, it often provided the best means of access to lower-lying 

 features. It was easier to walk along this bench and descend to 

 possible sites than to negotiate the talus slopes and rubble fields situated 

 at their foot. It was from this practice that most of the talus sites 

 were discovered. 



Sites on spurs and outliers (fig. 2-E) . — Artifacts are generally found 

 atop these features, and often occur in abundance (pi. 2:2). On the 

 higher, breccia-capped spurs and outliers, the material seems to be 

 indistinguishable from that of the bench sites. In fact, the inter- 

 vening cliff is sometimes quite insignificant, being only a few meters 

 in height, and a single spread of artifactual material covers both 

 bench and the upper portion of a spur. Flint or chert is often 

 locally available (from the breccias), and workshop debris is gener- 

 ally found mixed with the "finished" artifacts. These spurs fre- 

 quently protrude into the wadies in such a manner as to command an 



