The University of Michigan/University of Minnesota Expedition was in the field at Tel Kedesh from May 24 to July 17 1999. Our team varied in size from 21 in May and early June to 47 in later June and July. We conducted excavations in ten grids ranging from WE 3.2 in the north to CA 9.3 in the south. We recovered ca 2,300k of pottery and moved over 204,594 liters of dirt. The 1999 season revealed occupation phases from the Persian period to modern times. The most impressive remains were those of a large (ca 20,000 sq. ft.) late Hellenistic building at the SE end of the lower tel. We also completed the topographical map of the tel and its environs. We describe our finds by area below. The northernmost area excavated was the NE quadrant of grid square (No. 1) . There we uncovered two phases of architecture, both datable by sealed pottery to the Crusader period.
We put four small, 2x2 m probes (Nos. 2, 3, 4) in the area of the saddle between the upper and lower tels. Walls showing at surface and the topographical profile suggested that there might be a major Hellenistic gateway in this area. We put down two 2x2 meter probes in grid CD 5.8 on the south slope of the saddle, one in CD 8.7 on the north slope, and one in CD 7.8 at the low point in the saddle. Our excavation showed that the visible walls were relatively modern reuse of ancient materials. A modern utility pole under the latest architectural collapse in CD 5.8 dates that level to 1948. The latest material associated with the earliest constructions in CD 5.8 was Mamluk. The probe in CD 7.8 showed a similar deposit history. The probe in CD 8.7 yielded three successive ceramic horizons. The latest was modern with Ras el Fukra pottery as the latest datable material; the next was medieval with Mamluk material the latest finds. The earliest horizon was Late Byzantine/Early Islamic with significant amounts of restorable Byzantine period Havarit ware.
We opened a 5x7.5 m trench in grid WC 4.1 very near the center of the southern tel. Just below the modern surface our excavations revealed an apsidal stone structure built inside a rectangular stone frame. A monochromatic mosaic floor ran up the building on the east. Inside the apse were four heaps of stone oriented east west that are likely to be graves. We left these unexcavated and refilled the trench to protect them and the mosaic floor. The latest material from above the floor and around the heaps of stone was Byzantine. This and the architectural form of the apse in the rectangle lead us to postulate that the structure is a Byzantine mortuary chapel.
The remainder of our excavations areas were centered on the large late Hellenistic building on the south end of the south tel and neighboring houses. To the west of the large building we expanded the trench in which our 1997 probe had revealed a house abandoned in the mid-2nd c BCE (WB 3.1) The 199 excavations revealed a courtyard with cooking facilities to the east of the original room and another house to the north, slightly later in date, most likely late 2nd c BCE.
We opened four grid square in the large late Hellenistic building itself, one over the area where our 1998 magnetometry showed the NW corner of the building, (CB4.8, no. 7) and another where we expected the SE corner to fall. (CA 9.3, no. 10) We found the corners in both squares. The NW corner appeared ca 30 cm below the modern surface and was constructed of massive cut boulders. We also found walls of Byzantine date above the late Hellenistic building and walls of earlier Hellenistic date to the north of the building. Only a small segment of the interior of the late Hellenistic room lay in the area excavated. We left that at the level of fired mudbrick debris that we tentatively associate with the abandonment of the building after the 145 BCE battle between the Greeks and the Maccabbees near the site. The SE corner of the building was uncovered in trench CA 9.3. It was of similar construction to the NW corner and lay 70 cm below modern surface. Interestingly, it was founded on a massive earlier wall, which, from the latest material deposited against it appears to be Persian in date. Other finds (see below CB 2.4) indicate that there may have been a monumental Persian period building structure predeccesor to the large late Hellenistic structure.
With our two finals squares we explored the interior of the large Hellenistic building. In trench CB 2.4 we uncovered sections of four rooms of late Hellenistic date, one as large as 5.2x4.5 m and another only 3x3 m. A beaten earth/plaster floor 40 cm below modern surface in the NW room yielded a significant amount of mid 2nd c BCE pottery as the latest datable material in and under its makeup. Interestingly the walls of this room were built from large reused column drums placed on their sides. These indicate the existence of an earlier monumental building on the site, possibly contemporary with the Persian wall discovered not far to the south in trench CA 9.3.
Our second trench inside the building lay just to the east of the NW corner. (CB 4.7) This trench exposed a small slice of the NW corner room and all of a storeroom to the east. It was in this area that the most spectacular finds of the season were made. The storeroom had a cement/plaster floor ca 1 m below the modern surface with a deep, plaster lined settling basin along the east wall. We found relatively intact storage jars lined up against three walls of the room. There were two imported Rhodian wine amphoras, one Phoenician semi-fine baggy jar and 12 pithoi made of a local fine fabric. We also found a complete Phoenician semi-fine jug on the floor and fragments of two Campana A fishplates at the bottom of the basin. Preliminary readings of the stamped handles from the Rhodian amphora place them in late Period IV, giving a mid-2nd c BCE terminus post quem for the abandonment of the building.
The most surprising finds of the season came from the small slice of the corner room excavated to the west of the storeroom. Here we found a deposit of 12 complete unguentaria and amphoriskoi in the southern sector of the room. To the north of the amphoriskoi in an area ca 1x1 m we found over 1800 stamped clay bullae, of the kind used to seal and identify official documents written on papyri in the Hellenistic era. It is clear from this find that the room must have served as some sort of archival as well as pottery storage. We were able to clean only 230 of the bullae while in the field, as they were discovered in the last three days of the excavation. The cleaned sample shows a remarkable range of images from Greek realistic Aphrodite types to Phoenician schematic Tanits, portraits of Hellenistic monarchs and private individuals as well as some Persian and Egyptian types. The bullae are quite small averaging only 1-2 cm in height. Their backs show the impression of the papyri against which they were pressed. Many of the stamped images are small masterpieces of the engraver's art.
The Kedesh bullae hoard is one of 11 such Hellenistic collections extant. The largest is from Seleucia-on-the-Tigris (30,000) and smallest from Elephantini in Upper Egypt (35). Ours is the 4th largest of those found in situ and the 6th largest overall. Since we have excavated only about 20% of the archive room we can expect with some confidence to find considerably more bullae when we return next May. Most important, whatever the size, we will have the complete archive.
The finds from the 1999 season have led us to revise our original thoughts about the nature of the Hellenistic settlement at Kedesh. The scanty literary references to the site led us to expect a small farming village and/or garrison on the site in this era. The discovery of a massive public building with an archive and major storage facilities leads us to speculate that the site may have been a major administrative center of the region during the Hellenistic era and possibly the preceding Persian period. We will continue to explore these hypotheses in the next two seasons of excavation.