_Roman period_ Close to the ancient course of the Castra irrigation ditch, now dry and recently substituted by a channel with raised banks, excavations in trench A uncovered the remains of a single roomed building. Rectangular in plan (10.60 m x 7.20 m = circa 35 x 24 Roman feet) its walls were oriented east-west and north-south. At least two construction phases were identified: the earliest using stones, brick and mortar and the later phase using only brick and mortar. A further intervention in the trench identified the foundation constituted by concrete, which had been poured in liquid form into shuttering formed by wooden planks supported by posts. A number of these post holes were found. The flooring – resting directly on a natural stratum of gravel – was made up of large terracotta slabs or upside down winged tiles, of which two dumped examples were found. This particular building technique and the abundance of tiles, imbrices, slabs and bricks, (almost all stamped with Ti Nucula or variants), in the destruction layers, suggests that in the antique period the supply of materials was guaranteed by the presence of kilns in the area. On the basis of the pottery evidence the structure, interpreted as a storage facility and not as a dwelling, did not survive beyond the early imperial period. This was a service room belonging to a Roman villa destroyed during the 1970s when sand quarrying began. The fact that the building was on a perfect east-west alignment on its long side was of great interest. It therefore appeared independent with respect to the principal centuriation axes at Aquileia and corresponded with the alignment of the rustic villa at Pavia di Udine (in its mid 1st century B.C. phase) and the traces of centuriation identified in the area of Tricesimo, which recent research dates to within the first decades of the 1st century B.C. _Pre-Roman period_ Evidence for pre-Roman occupation came to light to the west of the Roman building in a number of trenches near the old course of the Castra irrigation ditch. This evidence comprised a surface for pottery working in the form of a levelled area of terracotta fragments probably pertaining to craft-working installations or small kilns and obliteration layers (with three pottery fragments one perhaps of Bronze Age date and two from the iron Age) cut by a series of Roman ploughed surfaces. Therefore, it seems that clay working and the production of terracotta were a tradition in the area over several centuries. There were numerous fragments of coarse and smooth impasto, perhaps interpretable as wall facing, and two fragments of truncated pyramid shaped weights of uncertain function which find parallels in the prehistoric excavations at Concordia.

Following the reopening of trench C, an east-west trench at a right angle to that of the previous year was opened. A particularly interesting situation emerged and was better defined with respect to 2009. The US 37 was seen to extend at the base of the known stratigraphic sequence. This was a surface comprising fragments of prehistoric baked clay and occasional fragments of Bronze/Iron Age pottery, dumped and compacted in a greyish clay-sand matrix. This was overlain by an occupation/abandonment layer containing scattered small, sharp gravel and baked clay and pottery fragments of the same date, resulting from the alteration of the layer below (US 38). This was probably a man-made bank along the course of an ancient river which, still in the prehistoric period, became a marsh. This was in fact demonstrated by the identification towards the west of a blackish brown layer with very occasional baked clay fragments (US 60), patches of which covered US 38, and deepened at first gradually and then suddenly to at least one metre, as shown by core sampling. The organisation of US 37 seemed to be connected with the gravel and sand substratum US 43, whose level descended notably in correspondence with the natural riverbed, along a front documented for at least 5 metres, extending from north-west to south-east. In the excavated area an ancient robber trench (US - 40) of an undefinable structure (kiln?), in phase with the basal layer, was partially uncovered. The cut contained various dumps, both primary and secondary, of baked clay fragments, ash and charcoal. Subsequently, with the area’s progressive abandonment, but still in the prehistoric period, further cuts destroyed and altered the pre-existing layers, only partially sparing the preceding stratigraphic situation. Two overlying levels of plough soil were identified: the first certainly Roman (US -48, 49), the second either Roman or later (US -51, 52). From the section it was clear that these cuts tended to deepen from west to east, marking a difference in height probably due to a preceding geomorphological situation. In the eastern part of the section it was seen that cut US 48 formed a wavy line from the top towards the bottom, the spacing more or less regular. The most convincing hypothesis is that this represents plough furrows, which on the basis of the discovery of one in plan cutting the gravel layer, ran north-west/south-east and were orientated to circa 320° North. Modern maps seem to show that the area of the Roman structure (found in 2009) is at 4.5 m a.s.l., whilst the area beyond the irrigation ditch lies at between 1.8 and 4.00 m a.s.l. This seems to indicate that already in antiquity occupation was concentrated on the highest part of the area, whilst the lower areas may have been subject to swamping caused by the presence of small water courses, as attested by the peat layer seen in some trenches.