After a third week of intensive surfaces collection the urban survey team supervised by Femke Martens (KU Leuven) completed the survey within the town. A last area of 14 sectors (0.56 ha) was covered to the northwest of the fortified area, west of the Doric Temple and just south of the middle Roman northern necropolis with its arcosolia tombs. The geophysical survey revealed the presence of a large building here, just north of the Hellenistic/late Roman fortification wall and parallel to it (see Survey, Geophysical, June 29-July 1). This building must have been laid out along the presumed northwest access road to the town center. As opposed to the eastern squares of the area, containing low densities of a range of finds (crustae, tesserae, metal, and glass), the western sectors within which the large building was detected by the geophysical survey contained very few finds. This is remarkable, since the surface visibility seemed to be roughly comparable in all squares. The processing of the functional and chronological data in relation to the results of earlier surveys in this area will bring more clarity in this matter. A particular point of interest for the chronological results of the survey in this area is the impact of the construction of the late fortification wall on the occupation outside of the walled area, which must have functioned as a refuge place for the surrounding population. The early fifth-century A.D. wall included many spolia, such as the blocks of the so-called Northwest Shrine, possibly a small temple, which was dated by L. Vandeput to the mid-second century A.D.

The 2004 season marked the completion of a sixth campaign of intensive urban survey. Two thirds of the total urban area enclosed within the necropolis of the town has been surveyed (ca. 11.5 ha by sampling and another ca. 10 ha with full coverage). The survey has already yielded a good picture of the chronological evolution of the occupation pattern as well as of the functional organization of the urban area. So far, the main indicators for the urban area's functional organization have been the architectural or sepulchral elements as well other non-ceramic finds on the surface. This year, however, it was decided to include also the ceramics into this functional research. Therefore, the survey team has spent six working days in the excavation house for the functional analysis of the collected pottery, in collaboration with Jeroen Poblome and Philip Bes. As such, the ceramics collected during the 2003 campaign in the western domestic area and the larger part of this year's pottery have been identified according to functionality. A classification was made distinguishing between pottery intended for agricultural production (production, transport, storage); kitchenware (intended for food preparation or cooking); table wares (for serving or consumption) and pottery for other purposes, such as cosmetics (e.g. unguentaria).

Finally, this week's work was concluded with a survey experiment, related to the comparability of the results of the 2000 intensive urban survey using the 10 x 10 m grid system and the results of the later surveys (2001-2004) using 20 x 20 grids. An experiment in 2002, in collaboration with archaeologist In Aé Delhaye and geologist Patrick Degryse, showed that walkers covering a smaller grid skimmed the surface more carefully, since they were physically closer to each other, but also risked a greater overlap of transects, resulting in a bias of densities of non-collected finds (such as building materials), which could be double counted. Within a larger grid there seemed to be less overlap of walker transects, which resulted in a less extensive coverage, but a more reliable density count.