The remains of the Roman circus were discovered in 2004 during archaeological investigations associated with the redevelopment of Colchester Garrison, south of Colchester’s walled town.<6>

This is the first Roman circus to be identified in Britain and is one of the larger (though narrow) examples known when compared to other circuses within the Roman Empire, being 448.2m long and 71.1-74.2m wide (excluding buttresses).

The Circus remains appear to be poorly preserved: all that survives of its walls are foundations, and many of these have been robbed out in the medieval (or later) period. Very little evidence of floors and other horizontally-bedded layers which might have provided stratified material were recovered during the excavation work. The foundations were constructed almost entirely of greensand, probably from Kent and were made of rubble set in mortar. A distinctive feature of this method of construction was the inclusion of many stone chippings, which must have been masons’ waste created as lumps of stone were prepared to make ashlar blocks for the wall faces. Part of the lowest course of the wall proper survived in situ at one point only (the middle of the three bastions in Site C2), where a row of small ashlar blocks formed part of the base of one of its faces.

Dating evidence for the circus is to date limited from the remains uncovered. The use of greensand in construction may be seen as an important indicator of the date of construction as previous excavations in the town centre have revealed that greensand was not widely used in Colchester until the late 1st or 2nd century. This is particularly clear from viewing the town walls which were built almost entirely of septaria and brick and are dated to c.AD 65-80. Recovery of the base of a black burnished ware bowl securely stratified from the gravel floor was dated to the c.AD 110/120 to 300. It is certainly possible that the gravel floor may have been replaced during the life of the circus but if it proves to be original, this find would provide further evidence supporting a 2nd century date for circus’ construction.

So far there is no clear confirmation of a terminal date for use of the building in Roman times. A possible indication of this is provided by the distribution of burials in the surrounding area. No burials have yet been noted within the footprint of the circus, despite the many which are known from the surrounds. Their absence may be a consequence of the circus remaining in use until the end of the Roman period, but the evidence is not conclusive. If the circus survived until the end of the Roman period, then to judge by the fate of other public buildings in the town such as the Temple of Claudius and the theatre, it is likely that it stood as a continuously degrading ruin until the Norman period. The circus may have survived even longer than that: a thin layer of demolition debris overlaying the gravel floor through the entryway on Site J East contained a single grey-ware body sherd, and its fabric, along with what appeared to be part of thumb-impressed strip applied to the external surface, points to a 13th or 14th century date. Although too much weight should not be placed on a single, rather unusual sherd, it does suggest that the circus may have stood as a visible ruin until at least the middle of the medieval period.<1><4>

It was originally believed that the starting gates of the circus at Colchester were situated at the eastern end.<2> However, further investigations revealed that the starting gates were positioned at the western end of the Circus, within the area of the D-shaped lawn on the (south) east side of the Officers' Quarters of the Royal Artillery Barracks (now decommissioned).<1>

During the 2007 evaluation at Area B1b of the Colchester Garrison (St John’s Abbey), five trenches were specifically targeted to investigate the east end of the circus. Modern and post-medieval intrusions had cut away the cavea wall/foundations in some areas. One trench positioned within the circus racetrack area found greensand blocks lying within a wide feature which may be the robbed or collapsed remains of the eastern end of the spina. An interesting find from the evaluation was a small piece of red granite found residually in one of the trenches, which was originally thought to be associated with the circus, possibly as part of the obelisk that may have stood on the central point of the spina. Unfortunately, it seems more likely to be associated with the later abbey site, and may be a fragment of an internal architectural detail of the abbey church, possibly a tomb.<3>

A robbed Roman wall on an E to W alignment was uncovered during trenching works in Area CX1 (Flagstaff Road/Circular Road North) in May 2004. This is on the alignment of the (later confirmed) Roman circus and is dated to the late second to mid-late third century so is very likely to be directly associated. The wall appears to have been robbed in the post-Roman period and a second Roman or later demolished wall was also recorded which had been built using Roman material.<5>

Further investigation of the circus occurred during the redevelopment of Area B1b (previously known as Flagstaff House, now known as Arena Place) in 2015-17 (see ECC2911, ECC2912 and ECC2913). The evaluation trenches T1 and T2 (ECC2913) and excavation Areas A and B (ECC2912) were specifically targeted over the circus. The aim of most of these investigations was to establish maximum dig-depths for the development to preserve as much of this important monument in situ as possible. As a consequence, few of the features directly associated with the circus were excavated, aside from those in Area A which were located on a major service line and not possible to avoid. Approximately 17m of robber trench over the outer cavea wall was exposed during the current project, which included one buttress in Area B. Although excavation was limited, no in situ wall foundations were apparent. In contrast, only 7m of robber trench over the inner cavea wall was exposed but a small section of in situ wall foundation to the east of Block H had survived. At 0.7m wide and made of fragments of greensand stone set in a yellow mortar, this surviving section of inner cavea is comparable to other surviving sections elsewhere along the length of the circus. The large pit (AF166) located in the centre of the barrier was of similar shape and dimensions to the probable water tanks/cisterns at the western end. All three pits included fragments of painted opus signinum in the backfill, perhaps used as a waterproof lining. Although no iron water collars were found in Area A, a gully (AF164) did appear to drain into the pit providing further evidence for its use as a water tank. A robber trench (AF165) on the northern edge of Area A is probably part of the perimeter wall of the central barrier; it was, however, much larger than seen elsewhere. Two small patches of buried topsoil identified immediately to the south of the inner cavea in excavation Area B (BL52) and during monitoring in Block H (WBL199) area probably parts of the arena surface. A medieval pottery sherd of late 11th- to 12th-century date and fragments of peg-tile were recovered from robber trenches and layers of demolition, along with Roman finds including pottery, ceramic building material and building stone. This would indicate that the final phase of robbing of the circus structure dated to at least the 13th/14th century.<7>