Area 1. A square (4 × 4 m) was opened, revealing natural bedrock, overlaid with sediment that was 0.3 m thick to the south and 0.1 m to the north. Only scant modern debris was found.


Area 2 (Fig. 3). A bedrock-hewn cave was uncovered and partially explored. An opening was breached in its ceiling (1.2 × 1.4 m) to access it. Two concrete piles had disturbed the cave, in the northeast and in the south. The latter prevented the extension of the excavation in this direction, thus precluding the restoration of the cave’s plan. The fill layer, which was mainly a fine and wet sediment (L1002), contemporary with the construction works that damaged the cave, rested on concrete, covering the original floor of the cave. The original entrance to the cave was probably on the south side, as no other entries were detected and the slope ascended toward the south and southwest. On the basis of these observations, as well as elevations taken, it was concluded that the cave, which was devoid of finds, was hewn to the north, and then westward, resulting in the shape of a spiral around a central pillar.
The cave may have been a rock-hewn tomb, although its plan does not correspond to Roman-Byzantine tombs that had previously been exposed at Sajur. A possible comparison for a cave around a central pillar can be cited from Qedesh in the Upper Galilee, which was dated to Middle Bronze I and interpreted as a shrine (M. Tadmor, IEJ 28, 1978). The complete absence of finds at the Sajur cave could be either a result of the cave not being finished and therefore not used for burial, or a complete plundering of its contents. At any rate, serving another capacity rather than a tomb can not be ruled out for this cave.


Area 3 (Fig. 4). A probe (5 × 7 m) was dug by a backhoe, 11 m north of Area 1, and excavated manually. It consisted of a stratified fill whose bottom part contained finds ascribed to the Ottoman period, including a jar's base (Fig. 5:1) and two storage jar rims (Fig. 5:2, 3), the latter belonging to the Rashaya el-Fukhar ware, dating to the nineteenth–beginning of the twentieth centuries CE.
The upper layer of the fill (L1001), which consisted of demolition debris and modern finds, sealed building remains to the south. This layer was rather deep and rested on a fill of light beige sediment in the northern half of the probe, which contained residues of a hearth (L1004) that was set on natural bedrock. Finds associated with the hearth comprised handmade ware, dating to the Mamluk or the beginning of the Ottoman period (Fig. 5:4).
The south half of Area 3 consisted of wall remains that delimited the southeast corner of a room. The eastern wall (W100) was constructed from two faces of dry-laid lime fieldstones and preserved three courses high. At its northern end, the upper course consisted of a bonding limestone block, laid as a stretcher, which rested on two reused crosette cornerstone blocks that originally were, most probably, elements of a door casing. Wall 100 abutted Wall 101 on the south; its northern face could only be cleaned. It was built of dry-laid stones and preserved two courses high. The very poor remains of a third wall (W102), preserved a maximum of three courses high and built of dry-lime fieldstones for c. 0.5 m, were uncovered. It abutted the northern end of W100. Most of the area enclosed by W100, W101 and the western section of the probe was covered with the remains of a flagstone pavement (L1003), which rested on a bedding of small stones over a thin fill that was intended to level the natural bedrock slope (L1006; Fig. 6). The fill that was deposited in places where flagstones had been robbed (L1005) was excavated independently from the top floor level down to the stone bedding. It included the fragment of a Mamluk or Ottoman-glazed pottery (Fig. 5:5, probably green and yellow gouged ware) and the rim of an Iron Age jar (Fig. 5:6). Finally, after dismantling the pavement, the bedding and the fill below were excavated down to bedrock, yielding Iron Age pottery fragments, such as a rim of an Iron II bowl (Fig. 5:7), an Iron IB cooking-pot (Fig. 5:8), the rim of an Iron Age jug (Fig. 5:9) and a lamp fragment (Fig. 5:10). The fill (L1006) contained a single non-identifiable rim fragment and some body potsherds.


The Ottoman-period pottery finds discovered above Floor 1003 belonged to the occupation phase of the building. However, the finds below the paved floor indicate various phases of Iron Age occupation, notwithstanding the piece of a plastic comb that was undoubtedly intrusive. Local inhabitants said that an old building was still standing in the area until recently and was destroyed prior to the excavation.
Against the façade of a modern building nearby and dispersed around the excavation area were architectural elements that included column bases and shafts (Fig. 7).


The built remains uncovered at the site cannot be interpreted as belonging to the Iron Age. However, despite the limited excavated area, the presence of many finds at the bottom of the stratigraphic sequence is, without any doubt, evidence for an Iron Age occupation at Sajur. This appears to be the main result of our excavation and should be considered in any future excavations at the site.

Y. Aharoni’s excavation (License no. 10/1951).

Aharoni visited the site prior to the excavation, describing thirteen loculi, most of them robbed and empty, while some were still filled with sediment. Some of the Sajur inhabitants mentioned various objects extracted from the tomb, but none could be verified. The few potsherds in the tomb were dated to the Byzantine period.

Aharoni excavated the two burials that were still undisturbed. The excavation, which did not yield any finds, was a definite proof for the early plundering of the burial complex. A schematic plan of the cave and sections with several measurements were drawn, serving as a base for the redrafted plan (Fig. 8). The entrance to the tomb was in the southeast and opened onto a quadrangular room, equipped with a peripheral bench and surrounded with thirteen burials: four in the southwest, five in the northwest and four in the northeast. One large loculus held four burials (5–8), one of which had a quarried pillow (6). Three burials were of the arcosolia type (1, 5 and 8); a small rectangular niche was carved southeast of Burial 1.

The plan of the tomb associates it with Roman-Byzantine tombs in the area. It compares particularly well with the tomb excavated by E. Braun at the site, to which it is also very close.