The largest hillfort in the Chilterns, Ravensburgh Castle, lies on private land, hidden beneath a plantation of trees. It is southwest of Hexton, in the northwestern part of North Hertfordshire district and the county boundary with Bedfordshire follows the outer edge of its western ramparts. Before the trees were planted about 1908, it was a prominent and impressive site, surrounded on three sides by steep-sided valleys.
Francis Taverner wrote the first surviving description about 1640 in an unpublished history of Hexton, now in the British Library. He said that it had a treble rampart, although it is not clear what he meant: there are two banks and two ditches, except on the east, where there a single bank and ditch. Eighty-four years later, William Stukeley visited the site and drew it from the west, confirming that there are only two ramparts on that side. He thought that the earthwork was Roman in date.
Ravensburgh appears on the early nineteenth-century Ordnance Survey sketches made about 1802 and has always appeared on their published maps, and Robert Clutterbuck published a plan in 1817. Quarrying inside the earthworks at an unknown date made the interior uneven before the trees were planted. Three rides in the woodland existed by 1940 but they are now overgrown.
In November 1940, Percival Westell, curator of Letchworth Museum, began digging a series of trial pits at the site. He started with two by the northwest entrance, by a higher part of the rampart known as The Keep. A week later, he dug three alongside the northernmost of the rides. Finally, he dug a trench down to bedrock in March 1941 just inside the southeastern entrance; the chalk was found 1.4 m below the surface. Unfortunately, whatever records Westell may have kept have not survived, apart from his plan of the site. No finds from his excavations made their way to Letchworth Museum, either.
John Moss-Eccardt of Letchworth Museum began more excavations in 1964, assisted by James Dyer, a lecturer at Putteridgebury College of Education, and his students. They dug a trench across the western ramparts, establishing that the site had seen three separate periods of construction, between the Early Iron Age (about 850-400 BC) and the Late Iron Age (about 100 BC-AD43). James Dyer returned to the site in 1970 and again every year between 1972 and 1975. Since 2013, Ian Brown of the University of Oxford has been working on the site. He has all the finds, Moss-Eccardt’s and James Dyer’s records, and is organising surveys.
The first phase consisted of chalk rubble packed between parallel timber palisades, with a flat-bottomed ditch on the outside. Long after the ditch had silted up, a new V-shaped ditch was dug and the chalk piled on top of the degraded first rampart as well as outside the ditch to create a counterscarp bank. At the same time, a second entrance was added to the settlement, near the southeastern corner. The outer ditch on the west side was added later, although its date is not known.
Geophysical survey inside the southeastern entrance in 2015 showed traces of at least two roundhouses and a small enclosure with an inturned entrance. Another, outside the northwest entrance in 2018, revealed a hitherto unknown Bronze Age round barrow, long since levelled but recognisable from the ring ditch that formed a quarry for the mound material. Recent surveys by Lidar (a technique for recording often subtle changes in the land surface) show that as well as the enclosed area, there are outer earthworks to the east running away from the northeastern and southeastern corners of the hillfort. It is unknown if they are contemporary with it and, if so, with what phase.
Analysis of the finds is ongoing and only interim reports are so far available. Nevertheless, Ravensburgh is one of very few sites in Hertfordshire with pottery from the Middle Iron Age (about 400-100 BC) as well as both earlier and later material. Although James Dyer thought that most of the pottery dated from his first construction phase, there is a lot of material from the first century BC through to the Roman Conquest in AD 43. There are also Neolithic and Bronze Age finds, pointing to the long use of the hilltop.
Written by Keith Fitzpatrick-Matthews
In the early 1880s, a tenant farmer of William Ransom found pieces of ‘brick’ on the surface of a field close to Purwell Mill. When Ransom visited, he found Roman pottery, brick and tile. In November 1884, he began digging and found a concrete floor just 0.76 m below the surface. He soon discovered a furnace for heating an under-floor heating system that fed three rooms. Beyond these were more rooms and a corridor.
He recognised that he had found a Roman bathhouse and assumed that it was part of a villa, a large country house. Traces of burning,
animal bones and oyster shells covered all the floors. Ransom thought that this showed that ‘some semi-barbarous tribe took possession and dwelt there’. The coins he found in it run from Gallienus to Valentinian II (AD 253-392), while others found nearby included one of Septimius Severus (193-211).
Percival Westell, curator of Letchworth Museum, decided to investigate the site further in the autumn of 1921. His team found the site of Ransom’s excavation and located a tiled corridor about 2 m wide. The finds they made were like those found in 1884, but they decided that the burning was a result of the buildings being destroyed by fire. Walking north of the site of the excavation one afternoon, the team found Roman pottery several hundred metres away, suggesting the site of a second building.
The Cambridge University Committee on Aerial Photography’s aerial photographs from July 1960 show the walls of the building. They reveal that the walls do not extend far beyond those that Ransom found: what he discovered in 1884 was an isolated bathhouse and the main dwelling must have lain elsewhere. The second site discovered by Letchworth Museum in 1921 was perhaps the location of the dwelling. The North Hertfordshire Archaeological Society did fieldwalking north of the bathhouse site in November 1978 and found more worn Roman pottery. Whether this was in the same place as the 1921 discoveries is unclear.
Over the years, illegal metal detecting has stripped the site of metalwork. Some detectorists brought coins into Letchworth Museum, including a coin of Antoninus Pius dated 145 to 161. Mark Curteis, then Assistant Curator of Letchworth Museum, re-examined all the coin finds from the site and concluded that they fall into two groups. One shows activity in and around the bathhouse, which probably began around the year 200. By the 330s, it had gone into decline but continued into the 380s or 390s. The second group of coins derives from a hoard, which Ransom excavated but did not recognise as such (although Westell later did). It dates from the period of the breakaway Gallic Empire of Postumus and his successors (261-274) and was buried in the mid-270s. The person who buried it never came back for it because coins of this period ceased to be legal tender in the later 270s.
Councillor Sam Collins walked around the field early in 2021 and discovered more Roman tile. He contacted the museum and organised fieldwalking one Saturday in September. Most of the finds consisted of pieces of Roman roof tile. There was very little domestic pottery and virtually no oyster shell, both things common where people were living. If this had been a villa site, where was the rubbish left by the people who had lived there?
There is perhaps not a villa at Purwell at all. The bathhouse is low-lying and damp, as Ransom observed in 1884. There is a risk of flooding from the nearby Ninesprings, which suggests an alternative explanation: similar isolated bathhouses close to water have been interpreted as parts of a nymphaeum. This was a temple dedicated to a spring or river, and the bathhouses used water blessed by the nymphs who lived there. A second building, visible on aerial photographs, stood on the hill-spur overlooking the bathhouse and could have been the temple. The burning discovered in 1884 and 1921 could then have been the deliberate destruction of a pagan site by angry Christians in the fourth century, something that has been found on other temple sites.
Questioning received wisdom – the things we think we know to be ‘facts’ – is always good. It is how we make progress, improve knowledge, and increase our understanding of the past. Just because something has been accepted for years does not mean that we should regard it as written in stone. As we find out more about the past and our ideas change, so we should be prepared to alter our interpretations of old discoveries.
Written by Keith Fitzpatrick-Matthews
In 1475, the newly-formed Guild of Our Lady Saint Mary the Virgin received a licence from King Edward IV to hold a property for their meetings. 2 Bancroft (The Brotherhood) has long been identified as the community’s Guildhall. As this part of Bancroft was formerly called Golden (earlier Gilden) Square, referring to the medieval Guild, the identification is plausible. Details inside the building confirm the likelihood, as the upper floor was built as one large room open to the roof. This floor originally projected out to the east, in what is known as a jetty (the overhanging part of a timber-framed building). Painted glass in the windows shows that this was a high status building, but it was destroyed in 1815 for sash windows, themselves removed at some time after 1885. It is now (January 2023) home to part of Lloyds Bank, Wilbury Clinic and Phone Tech.
Until the twentieth century, decorated tiles stood at each end of the roof ridge. Before 1912, when the Victoria County History described the originals as ‘still remaining in one of the shops’, replacements had taken their place. A photograph in the Lawson Thompson scrapbooks held in the museum shows that one of them was broken, missing its head, arms and reins. The business was probably Passingham’s wine shop at 2 Bancroft, as Alfred Passingham donated the tile to the newly formed (and still unopened) Hitchin Museum in 1939.
The tile measures about 315 mm in length and 256 wide. From the lowest part to the top of the rider’s head is about 360 mm, making it taller than it is long. It is earthenware with traces of a green (copper) glaze, discoloured through centuries of standing out in English weather. The horse and rider were made and fired separately from the tile and held in place with strips of lead. The example in the museum is mounted into a tile that is rounded, while the other was in an angled tile. An illustration in Reginald Hine’s History of Hitchin shows the surviving horse and rider on the angled tile; it is a composite of the two. It does not show the lead plugs, making it appear as if the tile, horse and rider were all one piece of clay.
The details of the rider’s head in the drawing are also inaccurate. The surviving rider does not
wear a ‘cap’ like that in the drawing, but rather a lobster-tailed helmet, with a forward-projecting peak. These helmets, based on a Turkish çiçak type, became popular after 1600 and had fallen out of favour by about 1700. This detail gives us a good date for the tile, making it well over a century younger than the building.
According to Hine, George Lewin (1832-1896) used to try to hit the riders – whom he called Hengest and Horsa – with missiles from his catapult when he was young. Could it be that he managed to hit the broken example? Mr Lewin – who played football for Hitchin Town FC in the 1860s – became the town’s constable, so he may not have wanted to own up to it. Local folklore has the riders coming to life: again, according to Hine, they would ride along the ridge of the roof every night, although the writer (Keith Fitzpatrick-Matthews) remembers that his mother told him when he was a child that they came down from the roof and did a circuit of the town at midnight on New Year’s Eve. In those days (the 1960s), the copies were still there, but they have since vanished. Where are they now? And where is the broken original?
In 2005, someone got in touch with the then assistant curator of Hitchin Museum, Caroline, to say that they knew where the missing tile was. They claimed that it was in a pond in a garden in The Avenue and asked if she’d like to go and ‘fish’ for it. It was a cold February day, so she said it might be possible later in the year. Then the trail went cold as the person never got back to her. Whether it’s one of the replacement tiles, the broken one or a tall story, we don’t know. A diver once told the writer in all sincerity that he had discovered a sunken Viking ship in the River Dee at Chester, only to vanish without trace (letters and telephone calls went unanswered). I am now suspicious about unevidenced tales of underwater discoveries, perhaps wrongly. If you know anything, please get in touch with the museum!
The first roof tiles to show a horse and rider date from the Chinese Tang dynasty (AD 618-907) and are known as liuli wa (‘roof tile of glass’), referring to their shimmering glazes. The technique is known as sancai (‘three glazes’) as the artists used a minimum of three different coloured lead glazes, usually green, cream and amber, and sometimes blue. These highly decorated products decorated both private and public buildings and were popular on tombs of the wealthy. They were intended to scare off evil spirits as the roof of a building is where the world of the living meets the world of spirits in the air. Another Chinese tile design that travelled west along the Silk Road in the Middle Ages is one of three hares arranged in a circle, so that each shares the two ears at its centre.
Hitchin Museum used the tile as its logo until the 1990s, as it has an iconic place in the history of the town. Although some people believe it to be medieval and original to the building, the design did not reach Europe before 1500, and the style of the rider’s helmet is a sure sign that it was made after 1600. It is now on display in the reception area of North Hertfordshire Museum.