Rome & Harting

Harting’s first intentional archaeological excavation was undertaken in 1934 by a retired schoolmaster from Fareham, Hampshire, called W A Gilmour. While on a visit to Harting he had been told by the then Rector, A J Roberts, about a place on the Downs, close to the parish boundary, where Roman coins & Samian ware had been found. Having obtained the permission of the landowner, at that time Admiral Sir Herbert Meade-Fetherstonhaugh of Uppark, he began operations. It was not the most promising of starts – the site was running to scrub, and the numerous roots caused him to give up after only a brief exploration without finding any significant features. Undeterred by this disappointment, however, he moved several hundred yards to the north, and began again on another possible Roman site at the head of a small valley in the Downs called Bramshott Bottom. The rest of his career as an amateur archaeologist was to be spent digging this second site exhaustively.

W A Gilmour at Bramshott Bottom in 1937, in the days when people knew how to dress properly while excavating

Unbeknownst to Gilmour his activities had been closely watched by a local builder’s labourer named Horace Brightwell. Once Gilmour had abandoned his first site, Brightwell moved in and continued to dig. It is possible that nothing may have come of this piece of archaeological poaching, had Brightwell not found a rare Iron Age blue glass bead, which he took to show the Rev A J Roberts. Roberts then sought to get it published without reference to Gilmour, an almost unforgiveable sin in the world of archaeology, and thus started a bitter verbal war of ownership which was to continue until Gilmour’s death in 1944. Brightwell even went to the length of naming his house “Bead Cottage” in a pointed statement of defiance.

All this is a rather long-winded way of answering one of the questions raised by my previous blog concerning what happened in Harting after the abandonment of the hillfort on Torberry Hill sometime around 100BC. For what Brightwell had found was the first clear evidence in Harting of a Late Iron Age artefact in close association with Roman finds. Gilmour was to go on to find many more, with his site at Bramshott Bottom producing a significant number of Iron Age coins amongst those of the Roman period. And this picture is by no means unique, for the majority of the Roman sites now known in the parish follow the same pattern, with their earliest known artefacts belonging to the first century BC and continuing on well into the Roman period, and in some cases beyond. What we don’t know is whether these sites go even further back, with Middle Iron Age origins contemporary with Torberry Hill, and this uncertainty is because only the two sites mentioned above have been excavated and neither of these was ever properly reported on.

However, our knowledge of Roman Harting is not restricted to merely the activities of Gimour and Brightwell. One of the big advantages of the increase in trade and technology that came with the Late Iron Age and Roman period is that there was a concomitant increase in the amount of stuff that everyone had. This in turn results in an awful lot more rubbish being generated, which then finds its way into the ploughsoil of our fields today. This means that sites of this period are considerably easier to find than those of almost any era of the past, and our map of Roman Harting is therefore much more densely filled. But there are two restrictions on this – first, to find a site it has to have been ploughed or disturbed in some way and secondly someone who knows what they are looking for needs to have been there to look.

The map below shows all the currently known Late Iron Age/Roman sites in the parish, and their distribution perfectly reflects the results of the restrictions mentioned above. The collection of five sites in the south-east part of the parish are all known from finds made in ploughsoil, mole-hills or rabbit burrows and all were discovered many years ago when previous owners of the Uppark Estate had close links to archaeologically-minded Rectors, or their friends, who were, it seems, permitted to roam almost at will. Included amongst these are the sites Gilmour and Brightwell investigated. The four sites on the west side of the parish were all located by metal-detectorists, at least three a product of large-scale detecting “rallies”. But only some landowners will allow detectorists access, and so such intense seeking activity is patchy across the parish.

Map of Harting parish (outlined in red) showing the known Roman sites (red dots)

With the exception of Gilmour’s site at Bramshott Bottom, which would appear to be a small farmstead, it is not known exactly what type of Roman site these all are, but they are likely to be either other farmsteads or small villas (the difference between the two is rather imprecise anyway, largely being define nowadays by whether the site has a timber or stone principal dwelling). But the cluster of three sites in the centre of the parish, lying within the village of South Harting, all lay claim to being definately of the latter type, the villa, with varying degrees of probability.

Way back in the 19th century, the then Rector, Rev. Gordon, recorded that he had spotted a tessellated floor and substantial wall within the grounds of what was then the Rectory on North Lane, leading him to assert that this was the site of Harting’s Roman villa. But his claim has never been proven, and the site remains doubtful. Some time later, during grave-digging and again later during excavations for an extension, Roman pottery and tiles were found in the churchyard, suggesting that it was here that Harting’s villa may have once stood, but a small collection of finds a villa doth not make, therefore this too is to be treated with caution.

The third central site, however, relies on much firmer evidence. The land either side of New lane has long been known to contain Roman artefacts, ever since Brightwell, whose house “Bead Cottage” lay on the lane, dug rather surreptitiously in the field behind and found pottery, tile and iron objects, before being ordered to stop. But then, in the early 21st century, an aerial photograph taken for reasons that had nothing to do with archaeology, revealed the clear outline of the foundations of what is called a “winged corridor villa”, thereby establishing Harting’s first, and so far only, confirmed Roman villa. To date it has not been further investigated.

Harting’s only known Roman villa – at least the outline of its foundations (courtesy of Chichester District Council)

With the close of the Roman period, Harting enters its own archaeological “Dark Age”, with no known Saxon site lying within the parish. Light only begins to shine once more once we enter the medieval world of the Hussey family, centred on the village of South Harting.

A Tale of Two Hillforts

Harting is in the unusual position of having not just one, but two hillforts dating to the Iron Age. Although before anyone gets ideas about two warring tribes fighting it out on the Downs, it must be pointed out that they were not contemporary. The first one sits on top of Beacon Hill and dates to the Early Iron Age, whereas the second one is on Torberry Hill and dates to the Middle and Late Iron Age.

Map showing the two hillforts, with the modern parish boundary shown by a red line

Very little of Beacon Hillfort has been excavated and so it is hard to be too precise about exactly when it was constructed, what it was for or when it was abandoned. Two attempts have been made in the past to answer these questions, first by P. A. M. Keef in the late 1940’s and early 1950’s, who excavated various small trenches dotted around its perimeter and then by Owen Bedwin in the 1970’s, who put a trench in the hillforts south-east corner and had another look at the entrance which Keef had early identified. Altogether these trenches looked at only about 2% of the whole fort, so not much to go on.

To help with the dating, the excavations produced a handful of Early Iron Age pottery and two rather spectacular gold penannular rings (sometimes called “lock rings” or “tress rings” because it was thought, somewhat fancifully, that they may have been braided into the hair of fair young Iron Age maidens). The rings came from the base of the ditch just beside the hillfort’s only entrance, which lies on its western side approached by a path rising diagonally up the side of Beacon Hill. Somewhat confusingly for those wanting to date the origin of the hillfort, the lock rings fit into a period from about 1000BC to 800BC, whereas the pottery was thought to be much later, but more recent studies of the latter are beginning to close the gap, and so it is now looking like the hillfort may well have its origins in the very late Bronze Age/ Early Iron Age, so around 800BC.

The Harting Beacon lock rings, now held in the British Museum

So what was the hillfort for? There would appear to be two trains of thought – some feel that it is too large (10 hectares altogether), too feebly defended (just one ditch and bank around its circumference) and too thinly occupied, to be the site of a major settlement, and that it was perhaps instead just a large stock enclosure, or a place for the occasional big meeting or market. Others point to the evidence of the artefacts that have been recovered (the pottery, animal bones, spindle whorls, loom weight and quernstones) and of the post-holes that were found in the excavations, to suggest that their must have been some people actually living there, cooking, eating, making clothes, and storing and grinding corn. Ultimately only more digging is going to provide the answer, but I would lean more towards the second school.

The rampart of Beacon Hillfort, looking into its interior (a handy herd of deer providing the scale) from its south-west corner

The date at which the hillfort was abandoned is even less certain than its beginnings, what we do know is that Torberry Hillfort began its life in the 5th century BC, and it seems not unreasonable to suggest that life on the Beacon had come to an end before then. It is tempting to think that the people of the Beacon decided that time had come for a change and that it was they that moved along the Downs a bit, to set up shop once again on Torberry, but we really don’t know, and it is perfectly possible that there was quite a time lapse between the two forts.

Torberry Hillfort is quite a lot smaller than its predecessor and this has meant that a rather greater percentage of it has been explored by the efforts of the archaeologists who put trenches into in the 1940’s and 50’s. Remarkably it was only actually discovered after WWII (Beacon Hillfort had been identified well back in the 19th century) and its finder was a local builder and grave-digger called Horace Brightwell. He it was who dug the first trenches across its rampart and ditch, and proved that it was Iron Age in date. In the 1950’s he was succeeded by John Boyden (and a youthful Barry Cunliffe, later excavator of Fishbourne Roman Palace), who continued its exploration.

What they found was that the hillfort started in the simplest of ways, with just a single bank and ditch dug across the top of the hill, and dividing it into two halves at a slight pinch point in its natural contours. The western part of the hill thus became the “in” part, and the remainder, with its more gentle slope eastwards, the “out”. But this was clearly not enough for those living within it, for shortly afterwards they extended the rampart and ditch to surround the whole western end of the hilltop, defending themselves on all sides. The urge to expand then came over them again in the 3rd or 2nd century BC, when they took in the rest of the hilltop. The final phase came with the construction of a very grand east entrance, composed of a 26m long stone-walled funneled corridor leading to a massive gate. And in the case of Torberry, no one disputes that this type of hillfort was densely settled, similar to Danebury in Hampshire (another of Cunliffe’s sites).

Torberry Hillfort from the Downs, the mist hides all but the top of the hill, where the hillfort lies, the early fort took in the left hand end, the late the whole visible length

Like many of the Iron Age hillforts still going towards the close of the second millennium BC, Torberry seems to have come to a juddering halt around 100BC, for reasons that are not immediately obvious. Contrary to popular belief, they were not all stormed by invading Roman armies, but instead settlement style and location seems to have shifted before the Romans even turned up, with such forts being abandoned in favour of lower lying ground. For Harting, the successors to Torberry are yet to be found, except from a few isolated finds in the fields around. Our next burst of activity being firmly within the Roman period. Of which more anon!


Water is obviously a key resource at any point in history, and this blog will focus upon its influence upon the history of Harting. Water obviously matters in terms of domestic consumption, for drinking, washing, cooking etc., but it also was the prime source of power, particularly until the advent of electricity. For water powered mills, and then later iron furnaces.

That part of Harting south of the Downs was pretty much water free, and so here the only solution was either to import it, catch it from the rain, or dig down for it in very deep wells. This is one reason why settlement was pretty sparse in these areas. But north of the Downs there was a fairly plentiful supply, and in particular from two streams that flowed from the base of the chalk. One of these streams flowed through West Harting Manor, and the other largely through East Harting Manor, although the latter also began in the west.

The two streams as traced from the 1632 map of West Harting Manor. The eastern stream can be seen starting within the deer park (outlined in yellow) but soon flows into East Harting Manor and is lost to our 1632 map

The eastern stream starts within the deer park attached to the Manor House of West Harting, just to the south-west of the modern village of South Harting. Its source is a spring running out of a little combe just to the west of The Warren, the water now flowing via a small pond down into South Gardens just below the lowest of the existing ponds there. However the 1632 map shows the spring being immediately dammed into two larger ponds, separated by a path, which then flow down to a building at the end of the now missing fourth pond of South Gardens. It is this building that is likely to be a mill, and probably a mill whose predecessors date back to Saxon times. In the Domesday Book, the reference to Harting includes a grand total of 9 mills, of which this was almost certainly one. The current ponds in South Gardens, also probably spring fed from the highest pond, may well have developed slightly later as fish ponds – the watery equivalent to the living larder provided by the deer park. After leaving South Gardens, the stream crosses the Chichester road and thus enters into East Harting manor, and out of the range of our 1632 map. Its lower course will be discussed in a later blog.

The origins of the East Harting manor stream, showing the spring-fed ponds feeding the mill (the upper shapes shaded green) and the later fish ponds of South Gardens (the three lower green shapes).

The West Harting manor stream originates from a spring near Ditcham, again at the foot of the Downs. It flows under the Petersfield Road (B2146) and through what is now Torberry Farm. From there to the River Rother, where it ends, it has an eventful journey, being dammed at least four times for various purposes.

Nowadays the highest dam is that at Hurst Farm, where a mill still exists, although no longer in use. This mill does not appear on the 1632 map, but is there by the date of the 1st Ordnance Survey edition of 1808. Further down stream there was an earlier mill, confusingly called “New Mill” which ironically now no longer exists. It used to lie in Goose Green, in angle between the road to Ryefields and that to Manor Farm, with the mill building located at the eastern end of a small rectangular pond, and several other buildings forming a small hamlet to the west.

New Mill at Goose Green. North is to the right. The mill is the rather faded building at the right hand end of the pond, with two more structures above, presumably one being the miller’s house.

Further downstream again we find what is now called “Harting Pond”, a much larger body of water. There is no sign of a building on the 1632 map in the vicinity of the pond’s dam, and no writing to indicate what its original purpose was. The most likely explanation is that it was the site of another of Harting’s nine medieval mills, but without further evidence this remains uncertain.

Harting Pond, lying next to the West Harting to Durford road. It is a much larger pond than either that of Hurst Mill or New Mill, but its function is not given on this 1632 map

We are on firmer ground with the next pond, although this time, conversely, no pond survives, but a name does. For on the 1632 map the words “Furnace Pond” are written over the area between channels of the West Harting stream, just as it exits the modern parish and enters Rogate, to the north-west of Nyewood. As the name suggests, this was not the site of a mill, but rather of an iron furnace, or more properly an iron hammer. The buildings of the hammer probably lay at the far eastern edge of the pond, inside the parish of Rogate, with the water backing up all the way to Parlour Copse or even Down Park Farm. Iron hammers tend to work in partnership with iron furnaces, with the latter extracting the ore from the rock and the former hammering it to produce a more marketable iron product. The pair for this hammer lay to the north in what is still called Harting Combe, although it lies now in Rogate parish, for here the site lay over the Weald clay, in which the ore can be found, whereas the hammer is over the Gault Clay, which does not contain ore.

The site of “Furnace Pond” to the north-west of Nyewood, on the border of the modern parish. By the date of this 1632 map the furnace has ceased working

The hammer and furnace were in operation from about 1588-1608, a relatively short period, perhaps suggesting that they were not very profitable, certainly compared to the many others that lay further into the Weald to the east. If you want to see the site of a furnace that lasted much longer and survives much better, then North Park Furnace at Fernhurst is the place to go (they have an annual Open Day in September which is well worth a visit (

Shortly after the hammer, the West Harting stream empties into the Rother, just to the west of Habin bridge, by Mizzards. And so also ends this blog ,except to observe once again that water meant power in centuries past, and a power that you could just keep on exploiting, building more and more dams to control and use its flow to power mills and furnaces. And the power was not just in the water, for whoever owned the water could make significant sums of money as a result of selling the produce of those mills and furnaces.