Links

Roman circus

(This article is extracted from The Colchester Archaeologist, 18 (2005), which is on sale now.)

The archaeological excavations are being undertaken by the Colchester Archaeological Trust (P Crummy, B Holloway, D Shimmin). The project is being managed by RPS Planning, Transport and Design (R Masefield, K Whittaker, C Lequesne). The investigations are being carried out on behalf of Taylor Woodrow who are funding the work.

Discovery

The first sight of the circus was in May 2000 when a Roman foundation was observed by CAT in the sides of a narrow trench dug for electric cables for some new lighting around the football pitch in the north-west corner of Abbey Field. The find was carefully recorded at the time but its significance could not appreciated because only a tiny fragment of foundation was exposed, and odd exposures of unattributable foundations of this sort are not infrequent in and around the centre of Colchester.

The next exposure was in 2002 when another short stretch was investigated during early archaeological work in a paddock near the Cavalry Barracks. Various trenches had been dug by machine as part of a sampling process to look for significant archaeological remains that might need further investigation prior to the redevelopment of the area for housing. Again, too little could be uncovered to make much of it. However, as we shall see, full-scale excavation later on made all the difference.

In 2004, two areas close to Flagstaff House on Circular Road North were excavated in quick succession. As luck would have it, one site turned out to include part of the south side of the circus, and the other the north. The remains took the form a pair of parallel foundations of unequal width. In both cases, the wider of the two foundations had square/rectangular projections along one side showing that it had also been buttressed. In terms of plan, one pair of foundations was the mirror image of the other, so that the narrower walls were to the inside of the circus and the outer wall was buttressed externally.

Excavation in progress on part of the Roman circus (Site J East)

The stone used in all four foundations is interesting for various reasons. Firstly, it is greensand. Stone of this type does occur in other Roman buildings in Colchester, but it is not the commonest material. It came a poor second to septaria, so the presence of greensand in both pairs of foundations provides some support to the view that they are likely to belong to the same building. Furthermore, the use of greensand gives us dating evidence for the building. This material does not seem to have been widely used in the earliest days of the Roman town -- it does not appear for example in the town wall which is datable to circa AD 65-80. Thus its appearance in the four foundations suggests that the building is unlikely to date to before the late 1^st century. Finally, much of the greensand was in the form of little chips and splinters showing it to be masons' waste, left over from making neatly-squared-off blocks of stone. These chips therefore show that the building, like the town wall, was probably of an ashlar construction (coursed inner and outer faces with a rubble and mortar core). The faces would probably have been made of perhaps four courses of greensand blocks alternating with four courses of brick.

Thus the two pairs of foundations resembled each other in composition and plan, making it reasonable to speculate that they were parts of the same structure. At that stage it was hard to imagine what structure that might be. Various possibilities were considered such as a temple precinct, but none of them seemed particularly convincing.

Then followed the excavation of some areas in the vicinity of the Cavalry Barracks. Most of it was concentrated in places where the machine trenching had earlier revealed Roman burials, but there was one area which was designed mainly to investigate the foundation mentioned above in the paddock. Not just one foundation was uncovered, but two looking in composition and plan just like the two pairs found in the excavations near Flagstaff House to the east. The newly-uncovered foundations were uncovered over a distance of about 75 m and, when carefully plotted on an large-scale Ordnance Survey map, turned out to align exactly with the southern pair of foundations near Flagstaff House. The alignment was impressive, especially considering that the two areas were about 230 m apart. Again it appeared to be the same structure. But what a size! Apart from the weird shape, it would have been too big for a temple precinct and it was hard to rationalise in terms of a defensive wall. Then the penny dropped -- a Roman circus!

An examination of the plans of various circuses known abroad showed that the circus is unlikely to have been much longer than the distance we had already traced it over. Could we find an end? Fortunately it proved possible to squeeze a few very narrow slit trenches in between some trees just west of the most westerly point we had traced the circus in the riding paddock. Our luck was in again, because we found the outer wall just starting to curve inwards. It was the beginning of what must be the semicircular west end of the circus. This fortunate find left us with the convincing plan of a circus.

Plan of the Colchester circus

Roman circuses

Chariot racing was the most popular of the mass spectator sports in Roman times. People flocked in their thousands to watch famous charioteers risk life and limb hurtling on light-weight chariots pulled at breakneck speed by teams of horses of various numbers. A day at the races was not only a time for great thrills and potentially lethal spills, but a chance to meet up with friends and an opportunity to place some bets and go home the richer for it.

Many, perhaps most, Roman towns must have had a circus of some sort. It needn't have been much -- two posts in a flat field would have been enough. More important places though had purpose-made stone circuses where the sides of the arena were lined with permanent seating set in tiers. These structures ranged in size and complexity from the Circus Maximus in Rome which was capable of holding a staggering 150,000 people (at least) to the more modest, yet basically similar, structures such as the circus at Colchester. The differences lie not so much in the size of the arenas, but in the size of the seating areas. For example, the arena of the Circus Maximus measured c 79 x 580 m which, as we shall see, is not nearly that much bigger than Colchester's as the difference in capacity might suggest.

There is more to the design of stone circuses than might appear at first sight. They were carefully planned with two principal aims in mind. One was that each of the competitors should have an equal chance of winning regardless of the gate from which they started. The other was that each of the spectators should have as good as view as possible of the arena and everything that happened in it. Essentially there are four elements which made up the circus: the track, the seating area, the central barrier and the starting gates. At one end of the arena were the starting gates. There would have been twelve of these, each built in stone and fitted with an iron barrier which would have been raised mechanically to start off the race. Two turning posts (metae) were placed in the arena and an elaborate barrier (spina) constructed between the two. The barrier would have included monuments of various kinds and lap counters in the form of dolphins and eggs. Many of the grandest of the circuses not far from Egypt included obelisks relocated from temples from that country.

One of the turning posts was placed in the centre of the semicircle forming one end of the arena. The other one was located about two thirds along the length of the arena, about 150 m or so from the starting gates. A white line was drawn on the ground between this post and the arena wall to the right of the charioteers. Each charioteer would have wanted to be the first to reach the line, because lane discipline ceased at this point and he would have been free to pull in front of his competitors if he could. It would have been a dangerous moment if all the chariots in a twelve-chariot race reached the line at the same time. The near post would have caused a bottleneck at the end of that opening dash to reach the line so, to ease the congestion, the post was sited four or five metres into what should have been part of the return track. This is why in circuses generally the central barrier is always slightly out of alignment with the sides of the arena.

The gates were not set in a straight line at right angles to the sides of the arena, because this would have meant that the distance to the white line would have varied according to lane. To ensure that the race was fair, the row of starting gates was curved just enough to make this distance the same for all the competitors.

The seating would have been raised in tiers to give the spectators a good view. In most circuses, the seating was provided on a stone substructure consisting of short vaults set side by side at right angles to the arena. The roofs of the vaults tapered downwards towards the arena to create a ramped base for the seats. In more elaborate circuses, there would have been a colonnaded gallery (portico) at the rear of the seating. Being roofed, the passage would have provided spectators with shade in the sunshine and shelter in the rain.

Colchester circus

Plan of Roman Colchester showing the location of the circus, Roman theatre and the Temple of Claudius

Much of the plan of the Colchester circus can now be reconstructed. Major elements are still missing, but the plans of circuses elsewhere allow the gaps to be filled in to a degree. The main uncertainty now is its length, because the end with the starting gates is yet to be located. All that has been found so far is parts of the seating area. The position of the starting gates is unknown, and nothing has as yet been seen of the central barrier. At Colchester, the arena was at least 350 m long and 69 m wide. One end was semicircular in shape where the chariots turned 180degrees. The other end was squared off to accommodate the starting gates. Tiers of raised wooden benches surrounded the arena except across the end with the starting gates. There was no colonnaded gallery as occurs in some other circuses. The benches were supported on a mound of earth retained in place by a low wall on the inside and a taller, buttressed wall on the outside. The distance between the two walls was around 4.7 m, which to judge by well-preserved circuses abroad, would have been enough for four tiers of benches. On the assumption that five people would need between 2.0 and 2.3 m of bench, the capacity of the Colchester circus would appear to have been around 7,000-9,000 depending on its length. Although unusual in circuses, the use of an earth bank for the seating rather than stone vaulting is a technique already seen in Britain in its amphitheatres and two of its four theatres (Verulamium and Gosbecks in Colchester). The thin external wall strengthened by buttresses is also particularly reminiscent of the external wall for the seating (cavea) at the Gosbecks theatre.

Nearly ten per cent of the seating area has been excavated to date. Small parts of the arena have also been examined, but this presents us with a problem. The arena floor needed to be soft enough not to lame the horses, but not so soft that it would rut easily. It probably took the form of a layer of sand over a solid base. But no trace of a surface or any sort of base have as yet been found. The upper levels of the arena were destroyed by cultivation in the medieval or post-medieval period; this is apparent on site. By whether the destruction went deep enough to remove all traces of the base is questionable. It may be that the arena of the circus at Colchester was simply a dirt track covered on the day of the races with a layer of sand.

Lingering doubts

Identification of the foundations as being parts of a circus seems almost certain, but more work is needed to remove lingering doubts. There are a few big problems. The foundations, especially the inner one, seem to have been rather shallow for the likely height of the walls they were intended to support. No physical remains of the arena have as yet been found except topsoil under it. Finds in general are very limited in quantity, and major elements of the plan have as yet to be seen. And finally, too little has been seen of the curve to rule out the slight possibility that the starting gates were in fact at the west end and not the east despite the hint of a curve apparent in the slit trenches dug immediately west of the most recent excavation area (see above).

However, the use of geophysical survey techniques (magnetometery and resistivity by Tim Dennis of the University of Essex)) should resolve some of these problems in the near future. Prospects for locating the central barrier by this means are good. More of the seating area may also be traceably by these methods too. An investigation of the west end is likely to be especially important since it should confirm the semicircular shape.

Date

The building of the stone circus is likely to have been the upgrading of an earlier, more modest circus. Indeed, there is an earth-filled feature under the outer foundation on the Cavalry Barracks site (Site J East) which conceivably might be a construction trench for a timber predecessor. But this is yet to be examined, and it might simply prove to have been an earlier boundary ditch of some sort. The presence of a modest number of mid-1^st-century glass cups in the town with circus scenes hint at a popular interest in circuses before the stone one was built. The foundation of Roman Colchester in the AD 50s would be the most likely period for the construction of an early circus. The town was the centre of the Imperial Cult (emperor worship) as witnessed by the Temple of Claudius, the remains of which can be seen in the basement of Colchester Castle. Circuses were associated with festivals and ritual as well as racing and other sporting activities, and a connection between an early circus such as this and the establishment of the Imperial Cult is a possibility.

End of the circus

The dating evidence for the end of the circus is even scantier than for its construction. Pottery in 'robber trenches' indicate that parts of the foundations were dug up for their building materials in the 12th century. The Norman period was a time when there was much building work in stone and brick in the town (eg the castle, St John's Abbey, St Botolph's Priory, and some of the churches). The work was undertaken almost entirely with reused Roman building materials obtained by knocking down standing Roman ruins and digging up many of their foundations. The east end of the circus appears to coincide with the southern part of the precinct of St John's Abbey where there are irregularities in the plan of the Norman precinct wall. Since the abbey was founded c 1095, it is conceivable that the odd shape of the precinct somehow reflects parts of the ruined circus which were still standing at that time.

However there is another possibility. Excavations at Balkerne Lane and Middleborough (both extra-mural sites) have shown that the suburbs of the Roman town were demolished without replacement during the last quarter of the 3rd century AD. The circus might have met the same fate. Had this happened, we might expect to find 4th-century burials within the site of the circus. But no such burials have been found, and the evidence is ambivalent.

Future of the monument

In the short term, there will be more (but limited) investigations to clarify various key issues to do with the plan and purpose of the building. In the longer term, those parts of the monument which lie within the Taylor Woodrow development site are to be permanently protected. The surviving foundations (at least those parts which have been uncovered so far) are probably too delicate to survive long-term exposure, so they are to reburied for safe-keeping. It is early days yet in terms of planning, but Taylor Woodrow are excited about the find and have stated that the circus is to be incorporated into their new development in such a way that the remains will be protected and the site of the circus left as open space, at least as much of it as lies within their development. Details are yet to be decided, but the intention is that the site of the circus will become a permanent and visible feature of Colchester in years to come. Indeed, a mosaic, designed by Peter Froste and made by schoolchildren and students at the Philip Morant School in Colchester, is intended to be installed at the circus site.

Philip Crummy
(Colchester Archaeological Trust)