An archaeological watching brief was carried out by Colchester Archaeological Trust, from April to November 2008, during the digging of trenches for replacement gas mains in Queen Street and Culver Street East, Colchester. They extended for approximately 210m south-north in Queen Street and into the High Street, and for approximately 220m east-west in Culver Street East. The trenches were usually 0.45-0.6m wide and 1-1.2m deep, although this varied, notably where obstructions were encountered or where they were linked up to other gas mains. The trenches were dug by a mini-digger with a toothless bucket. <1> The top of the natural sand and gravel subsoil (L8) lay, in most places, below the level of the bottom of the trench, especially in Culver Street East. It was located in a few places in Queen Street, notably at the southern end at a depth of 0.8-1.2m below the modern ground-level and probably also further north outside 2a Queen Street at a depth of 1.7m. It was reached by augering at the northern end of Queen Street at approximately 2.05m below the modern ground-level. During machine-trenching across the east end of Culver Street East and under the pavement to the east of the Natural History Museum, a thick reddish-brown layer L14 was revealed only 0.5-0.6 m below the modern ground-level. This consisted mainly of fragments of burnt daub and was presumably the same deposit as that which had been recorded previously and interpreted as kiln debris (Hull 1958, 205). <2> Cleaning over L14 after machining revealed the remains of two daub-block walls, aligned E to W (F19, F20). The top of F19 was 0.82 m below the modern pavement-level. F20 lay approximately 8m to the north of F19. The top of F20 was 0.72m below the modern pavement-level (note: the upper part was removed during machining). An auger-hole was dug immediately to the south of F20. This suggested that the layer of Boudican destruction debris (L14) was approximately 1.45m deep (ie 0.5-1.95m below the modern pavement-level). This appeared to seal a thin burnt floor deposit about 0.1 m thick, which in turn sealed natural sand (L8). It was unclear whether the burnt daub-block walls (F19, F20) belonged to the same building. Their method of construction was similar to that used in barrack-blocks within the fortress. They were, therefore, probably of military origin, built as part of the fortress annexe, and were subsequently re-used in the early colonia. Elsewhere, pre-Boudican remains were not often reached in the trenches. A layer of probable gravel metalling (L18) 0.15m thick, which was sealed by Boudican destruction debris (L14), was observed outside 2-2a Queen Street. It lay 1.36-1.51m below the modern pavement-level, at a point where the trench had been dug deeper than usual in order to avoid an obstruction. The metalling contained a few fragments of Roman brick/tile, and was traced for at least 3m. It possibly formed part of an east-west street between Insulas 30 and 38 in the early town. Deposits of reddish-brown Boudican destruction debris (L14) were encountered in a number of places along the trenches. L14 was particularly marked at the northern end of Queen Street and the eastern end of Culver Street East. To the east of the Natural History Museum, it lay only 0.5-0.6m below the modern ground-level (Sx 1 & Sx 2), and augering suggested that it was 1.45 m thick. It contained many fragments, some quite large, of burnt daub blocks. There were also moderate amounts of Roman roof tile and brick fragments, as well as sparse opus signinum fragments. Only the upper part of this deposit was visible in the trench, but much of this did not seem to have been extensively re-worked or redeposited in the post-Boudican period. This suggests that the lower parts of the layer may survive in situ. Further to the south and west, the layer of Boudican destruction debris (L14) thinned out and was not so close to the modern ground-level. Layers of Boudican destruction debris (L14) were also found in places further south in Queen Street, although not so extensively, and the burnt daub looked more fragmented and redeposited. Outside the eastern side of 1 Culver Street East (Prezzo) and the northern part of 2 Queen Street (Queen’s House), the top of L14 was uncovered near the bottom of the trench, 0.8-1.03m below the modern pavement-level, and was at least 0.2m thick. Post-Boudican street metalling was recorded in several places in or near the bottom of the trenches. A disturbed layer of gravel metalling or make-up (L5) was observed at the south end of Queen Street, immediately north of Short Wyre Street inside the walled town. This was up to 0.2m thick and was reached at only 0.5-0.6m below the modern street-level. It sealed natural sand (L8) and was probably part of the Roman north-south street leading down to the South or St Botolph's Gate from inside the walled town. Further north, a layer of gravel metalling (L16), 0.12-0.15m thick, was observed outside 6 Queen Street. This lay 0.55m below the modern pavement-level and was clearly stratigraphically later than the layer of Boudican destruction debris (L14). The gravel was only traced for 2.2m in the west section, but was probably more extensive as this part of the trench had been much disturbed by later activity. The gravel layer lies on the projected line of the Roman street at the south end of Insula 30, separating the latter from Insula 38b (CAR 6, 15, fig 2.9). <3> Outside 2 Queen Street (Queen’s House), a hard gravel layer (L19) was uncovered in the bottom of the trench. It was traced for approximately 9m and lay 1.1m below the modern pavement-level. The layer contained some large gravels and sparse fragments of Roman brick/tile and mortar, and was sealed by ‘dark earth’ (L4). The soil conditions in this part of the trench were damp. Several well-preserved wooden stakes or piles cut the upper surface of L19, projecting just above the level of the gravel. One example was recovered; it was 0.5m long. It was unclear if they were contemporary with L19, although a medieval or later date for the stakes seems more likely. A layer of gravel metalling (L24) was observed in the bottom of the trench outside 34/36 Culver Street East and the west end of the car-park of 66 High Street. It lay approximately 1.1m below the modern street-level. The metalling probably belonged to a north-south street sub-dividing Insula 30. Under the pavement on the west side of Queen Street, outside the east wall of 1 Culver Street East, the remains of a Roman drain (F12) extended obliquely across the bottom of the trench. It was traced for approximately 3m and was aligned NNW-SSE. The sides and base of F12 were constructed of Roman brick set in opus signinum. The bottom of the channel lay 1.16m below the modern pavement-level. Immediately to the south of the drain F12, the remains of a large unrobbed foundation (F11) were uncovered in the bottom of the trench. It was constructed of septaria fragments set in a hard, pale brown mortar. It was traced for 3.4m in the bottom of the trench and was cut into redeposited Boudican destruction debris (L14). It lay 1.05-1.25m below the modern pavement-level. It seemed to form the corner of a building near the southeastern corner of Insula 30. Outside the northern end of 6 Queen Street, there was a well preserved foundation (F10). It was observed in the western section of the trench, and was 0.95m wide and extended east-west. It was constructed of septaria blocks in a pale brown mortar, and lay 0.6m below the modern pavement-level. The foundation probably supported a wall or colonnade along the south side of Insula 30. Elsewhere, there were robbed out Roman foundations and Roman deposits/layers, indicative of Roman buildings, floors and surfaces. An opus signinum floor (which probably formed the base for a floor suchas a tessellated pavement) from a building in the south-eastern corner of Insula 29 was also recorded in Culver Street East, 0.95m bgl. Over much of the length of the trench, the Roman deposits were sealed by post-Roman topsoil or dark earth (L4; CAR 3, 92). <4> This dark greyish-brown deposit varied in depth, but was typically about 0.5m thick. This was the case towards the southern end of Queen Street, where it was reached roughly 0.45m below the modern ground-level. Gravel spreads were observed in a limited number of places within L4 - exposures of probable post-Roman gravel metalling. Several probable post-Roman robber trenches for Roman foundations were observed in the trenches, including F6, F7 and F9 in Queen Street. On the western side of St Nicholas Street, by the south-eastern corner of the graveyard wall, some unstratified human bone (finds no 39) was recovered from the trench. They probably came from the dark earth (L4) and belonged to a burial in the graveyard of the now-demolished St Nicholas' Church. Slightly to the north of this, an arrangement of several re-used Roman brick/tile fragments, presumably a feature, was observed within the dark earth (L4), 0.8m below the modern street-level.