New Light on Late Antique Pella: Sydney University Excavations in Area XXXV, 1997 1 Kate da Costa, Margaret O’Hea, Lachlan Mairs and Rachel Sparks. In January to February 1997, six trenches were laid out in a new area of excavations at Pella, Area XXXV, in the saddle between the main tell and Jabel Abu al-Khas (fig. 1). The four main trenches were grouped just inside the Department of Antiquities site fence, at the curve of the modern road, astride an old track down to the springs in the wadi. Surface wall lines appeared to confirm Robert H. Smith’s suggestion that a Roman period road had run from the Parvis, north in the saddle, towards the modern road (Smith & Day 1989: 6). The extant walls, and three apparent street paving blocks, suggested a bi- or tri-partite gate, in the manner, though certainly not as grand, as the South Gate at Jarash. In the main trenches (A-D) excavation soon revealed that the extant wall lines were mainly those of a courtyard house dating to the Byzantine and early Umayyad periods (fig. 2, phase 3), the three paving stones forming the top steps up from the courtyard to the street. Two other earlier phases and an episode of substantial filling were discovered elsewhere in the area. It is these earlier sequences which form the bulk of the description in this report. The entire area was affected by years of modern foot and herding traffic, as this was the major accessway for the local Bedouin shepherds to the springs of Wadi Jirm. It is also partly the track used by motor vehicles to access the tell. The natural topography of the area encourages erosion during rainy periods. However, drawing on experience with excavating historical sites in Australia, where the total depth of deposit can be as little as 20 cm and also heavily contaminated, we endeavoured to isolate those areas which had been somehow protected from contamination and were consequently able to recover clean sequences of artefacts to associate with the architectural phases. The solidly built walls of phases 1b and 3 seemed to have protected the interiors of the rooms from traffic, and it was mainly the south part of trench C which was the most cut-about.