Roman Road Winchester to London

Dark Circles beside the Roman Roads out of Chawton Park Wood

We have been excavating the Roman Road features on the western edge of Chawton Park Wood since 2006. It is believed to be the Roman Road from Winchester to London, which is also part of a Roman Centuriation where a Roman estate is provided with a grid of parallel and right angle Roman estate Roads (Refs 1 & 2). In the white field to the north of our excavation sites, shown S on the air photo map below, are a substantial number of Dark Circles. These are seen on light photo copies of Ref 3 air photo which under a magnifying glass appear to show ditches. In normal darkness photo copies they show as Dark Circles with diameters in excess of 20m, with more of them than shown below, including all the white fields to the north of our excavations.

There are also a large number of such circles on Ref 4 air photo, a mile or two to the SW of the above circles. One group is in the white fields to the north of Old Down Wood (where we excavated on a Centuriation cross-road) SU65 33. This is on high ground which is sloping to valleys to the north and south. The other group is to the east of Hawthorn SU68 33. These are largely in grey fields on the air photo but also in white fields over an area 1 mile in diameter. Clearly these features are common hereabouts.

In the above air photo map the summit of the downs is in the grey field to the west of S, at S the land is sloping down slightly to the east. In the white field to the north of S a valley is opening up running down to the NE approximately along the field boundary in the middle of the field. This valley has become quite deep by the eastern boundary of the white field. The Dark Circles are generally only clearly visible after the field has been cultivated. Then it is seen that the soil is generally red in colour, but over a Dark Circle it is the normal Dark Grey of Top Soil. Trowelling down does not find a bottom to this Dark Grey Top Soil. It is not superficial. There is no sharp dividing line between the Red and Dark Grey soils. Picture of Dark Circle

The farmer thinks these Dark Circles may be due to the fields being reclaimed from Chawton Park Wood within the last few hundred years, and result from burning waste timber from the felled trees. We have had problems with the stratigraphy in excavations around the Roman features in the grey field we do not find the normal pattern of Top Soil and subsoil instead we find the soils running very deep well below where we expect the natural clay to start. This seems to suggest some deep features have been dug and filled in before the time that soils turn into subsoil. This may mean we have some Dark Age features constructed over the Roman features and filled in when these fields were formed out of the Wood. Due to the difficulty of excavating at such depths we need to learn more about the stratigraphy of these fields to investigate this problem further.

The other possibilities for these Dark Circles: Bronze or Iron Age Round Houses this would be quite a big town; or Dew Ponds. The likely stratigraphy for these possibilities is levelled house platforms with occupational layers and finds, or ponds dug into the clay for Dew Ponds; and no change in the natural stratigraphy for bonfires. Some dating evidence may come from the presence of subsoil silting up levelled platforms or ponds if these features are older than medieval. Bonfires can be expected to produce a high positive magnetic anomaly as the natural is clay which will be fired to pottery where fragments occur in the soils.

The possibility of these being water features was of interest because a major Roman water feature was found in the excavation of one of the Centuration roads (Ref 5). Also the central of the three Roman features shown on the above air photo map is interpreted as a Roman Water Feature (Refs 6 & 10). So a trench was cut across the Dark Circle near the SE corner of the white field north of S on the above air photo map (there is in fact another smaller one right in the corner which can be seen with a magnifying glass).

The Composite Section shown above is along the E-W archaeological grid at 114.5N, which means the trench is parallel to the Roman features in the Dark Field. It was estimated to be across the centre of the Dark Circle. Trenches 1-4 were cut in 2013, with Trench 4 extending out of the Dark Grey Top Soil and into the red material occurring in most of the field except the Dark Circles. We can only excavate in these fields after harvest which can be late and before the new crop is sown. This tends to be every two years. In 2013 we came into a ditch in Trench 1, which indicated a further trench was needed, but in 2015 the availability of the field and manpower did not coincide, and Trench A was cut in 2017. The land is sloping increasingly to the north the maximum slope is near the E-W grid line.

Basically a thick layer of Top Soil rests on the natural Clay, with this Top Soil changing from Red to Dark Grey in western half of Trench 4. In Trenches 3 & A there are some thin lenses of subsoil on the Clay surface, and some thinner ones around the boundary of Trenches 1 & 2. In woods we usually get only a few cms of Top Soil and 20-25cms of subsoil. In worked fields we usually get this same thickness of subsoil, with 25-30cms of Top Soil. Our interpretation for what we have here comes from the likely relative recent recovery of these fields from Chawton Park Wood. The plough has combined the thin Top Soil and most of the subsoil into one layer leaving only thin lenses of the original subsoil in deeper folds of the Clay. Subsoil takes about 1000 years to form time has not yet elapsed for a subsoil layer to re-form. This information will be important for any further operations in the Dark Field.


In Trench 1 a ditch opened up, with and increased depth of Top Soil up to 10cm, and subsoil up to 20cms thick. There is a line on the clay on the Section photographs, shown dotted on the Composite Section, which may represent the original cut of the ditch which would be half a meter deep from the present ground surface. We have a lot of experience of Roman Road side ditches, where we usually can make out on the Section photographs silt lines, and re-cut lines. There are no silt lines visible for this ditch, and apart from the line in the clay mentioned above, no re-cut lines can be seen. This suggests that once dug, this ditch was not left open but re-filled, probably with soil but not clay.

As mentioned, the above air photo map indicates these features contain circular or semi-circular ditches. It looks as if they are a form of Dew Pond, where Ref 5 describes one source of water as seeping down a slope near the summit of the downs on an impermeable layer such as clay, to collect in a pit or pond dug into this layer. These Dark Circles are near the summit of the downs on sloping surfaces over clay, and water would collect in these ditches. As in the last paragraph, it is likely these ditches were re-filled with earth, so the possibility of a second source of water from dew in summer nights was not used (this topic is subject to experiments at excavations on high ground). So the water collected would be stored in the earth of the ditches, and seep out to the earth surrounding the ditches and stored there also.

Plan at the end of this report shows the extent of the Dark Top Soil as the outer dashed black line, plotted on the archaeological grid. The black crosses are points determined as just on red soil. The 1m wide trenches opened are shown in black, with Trench A just in the Dark Top Soil at the East end, and Trench 4 with its West end in the red material. The Ditch found in Trench 1 on the boundary with Trench A is shown in black. On the light photocopy of the Ref 3, 6": 1 mile air photo, the ditch is shown U shaped in the process of putting this on the web it has become a V shaped ditch. A scale model of the U shaped ditch was placed on the Plan to fit its occurrence in Trench 1 this is shown as the U shaped dashed black lines.


A magnetometary exercise was undertaken with our metal detector, the results are shown in Green on the Plan below. The Ground Balance over the Ditch was found to be 4 to 4.2 (no change in Signal from high in air to on the ground). The Threshold was set at 10 on the meter, which enables negative effects to be measured (but cannot be shown on the Plan below). If the Discriminator is set below the Ground Balance, a greater Signal is recorded on the meter, which is taken as the measurement of the magnetic effect of the ground. The greater amount of subsoil and especially Top Soil in the Ditch can be expected to give a greater magnetic effect and so a greater recorded Signal on the meter. Four traverses are shown on the Plan which were in the Dark Circle. It was not ascertained where the Ditch is liable to run. Each E-W traverse started on the N-S Base Line, and the Discriminator adjusted to give 10 on the meter (the Threshold). High Signals could be tested on the Discriminator if they persisted much beyond 4.5 they are liable to be metal e.g. Iron goes our between 5 & 6. Metal Signals were removed. Traverses at 108N and 119N had the Discriminator at 3.5, the other two at 3.9. The 40cms Detector Head was moved over a tape measure on the ground and the operator shouted out the Signal readings, especially where there was a significant max or min. The recorder wrote down the Signal reading together with the Head's position on the tape. The Signal recordings are plotted in Green over the traverse in the E-W direction. The Meter Scale of 50 above the Threshold is shown in the top right hand corner of the Plan below the maximum possible reading is 70. If these Dark Circles resulted from bonfires burning waste wood then we would expect the whole of the area of the bonfire to give a Signal of 70 clearly this is not occurring.

The Magnetometary results broadly support the position of the Ditch, though only two traverses cross it. Where a traverse goes near the Ditch an enhanced Signal can be expected. The Ditch is not likely to go further than 25W, and unlikely to go outside the Dark Circle where it would cross traverse lines. Its southern limb may go a few more meters W than its northern limb given by the dotted southern extension.

We can then see that the Dark Topsoil type material extends some 5m E (downhill) of the Ditch, and also N & S of the Ditch except where the Ditch comes to an end where this distance contracts to some 3m. It also extends the same 5m W (uphill) of the Ditch, especially if the Magnetometary suggests the southern limb of the Ditch extends a few meters more W to give the lopsided shape of the Dark Circle.

What we are seeing today is a modification of this Dark Circle by ploughing after the land was re-claimed from Chawton Park Wood one or two hundred years ago. The Composite Section indicates that most of the subsoil has been removed and mixed with a thick layer of regular Top Soil. Consequently we must assume before this reclamation there was a thick layer of regular Top Soil on top of the usual subsoil layer we find in this wood of 20-25cms thick. This means that a Dark Circle would have stood some 25-30cms higher than the rest of the field because in the Wood we find only 5cms of Top Soil not the 25cms of Top Soil we find in this Dark Circle. A further process will have been going on during the hundred years or more of cultivation the height of the Dark Circle will have been reduced by scattering of the Top Soil. Perhaps now they are 10cms higher. As their height was reduced so the plough would start cutting into the subsoil now largely removing most of it.

Water feature

We thus have a model of the original form of this water feature. A semi-circular ditch, 1.5m wide, 7m in diameter, cut 40cms into the clay, is able to increase the fertility of the ground to produce 25-30cms of regular Top Soil (whereas Woods only support 5cms), resting on 20-25cms of subsoil above the clay natural. Not only that but presumably these Dark Circles had remained unused by man during much of recent history, yet they have continued to function and maintain the thick layer of regular Top Soil we find today.

The main dating evidence comes from the Ditch filled with subsoil in Trench 1 which puts it before about AD 1000. We generally find that Roman levels are towards the top of the subsoil. Otherwise it is extremely difficult to date a ditch. We concluded above that this one had been backfilled with soil when cut out of the clay so finds in it reflect the finds in the back-filled soil, not the date of cutting. Charcoal finds occurred in the subsoil lenses but not in fact in this ditch. Finds at the base of the subsoil may well pre-date man.

While the subsoil dating could extend to early Saxon, Dark Age, and Iron Age, in fact the most likely dating is Roman. Romans had superior water technology, which we do not yet understand in say North Africa. We have come across a swimming pool sized example in a previous dig in Ref 5. Darker prints of the above air photo map show the white fields north of the excavation site to be filled with these Dark Circles. This is a large scale production site and others were mentioned on the Ref 4 photo. Large scale production sites are only known outside the Roman period from the Industrial Revolution this rules out Saxon or Iron Age as dating for these Dark Circles.

Our experience of excavating in these fields is that the soils and clay are moist in May, but by August (unless the summer is unusually wet) the clay is like concrete, and all moisture has been sucked out of the soils. We have concluded that water would be stored in the clay part of the ditch, which we estimate from the Plan below would amount to near 13m, or near 3,000 gallon of max wet earth. Irrigation assistance is required for the high summer months of June, July and August 1,000 gallons a month would seem likely to be sufficient for a Circle of this size. This suggests that these Circles are for supplying water dependant fruit & vegetables for the Roman city of Winchester.

For we must remember that we are on a Roman Centuriation estate the transport links are visible in the field to the south including the central one which is interpreted as a water feature. These estates might be constructed by the Imperial Roman Government (often for pensions), or they may be constructed by a City Government. A possible model was to construct the parallel and right angle estate Roads, which then made it easier to let out parts of the land. The southern of the three features on the above air photo map it thought to be the Limites or estate boundary, it being a Road leading directly the modern A31 E-W out of Winchester. Transport links for the produce and supplies are close at hand (Refs 7 & 8)


Thanks are due to Mr John Cray for permission to excavate on his land; and to the 52 man days contributed by the diggers. Richard Whaley

Map Caption The Air Photo Map made from Ref 3 Air Photo 41/63 NE from Chawton Park Wood to Soldridge (SW of the Photo).

Composite Section Caption Trench cut through the Dark Circle near SE corner of the light field in the above air photo map which shows as a ditch. The height scale is below Datum 1.

Landscape Printing Option

If you want a full size Landscape Print of this Composite Section: Check your margins 19mm and set to Landscape (File, Print Set-up, Landscape) Click After printing, re-set to Portrait.

Caption for the Dark Circle Plan below The Black Dashed outer line is the extent of the Dark Circle, the crosses are just on the Red soil. The Trenches A and 1 to 4 are shown in Black. The inferred position of the U shaped ditch is shown in black dashed lines, its position in Trench 1 is shown in black lines at SU6682 3657. Magnetometary readings are shown in Green, the Meter reading for 50 above threshold is shown in the top right hand corner.


1. NEHHAS Journal 3 2007-14
2. Note on Centuriation, NEHHAS NL No.91 Spring 2004, Journal 3 No.10 2007
3. RAF air photo survey 1947-9, scaled to 6": 1 mile by OS, Hants Sheet 41/63 NE
4. RAF air photo survey 1947-9, scaled to 6": 1 mile by OS, Hants Sheet 41/63 SE
5. Richard Whaley, Did the Romans have advanced Dew Ponds?, NEHHAS FAB e News 8, Autumn 2013
6. Excavations on the Central Feature FAB e News Nos. 1 & 3, 2010-1
7. Excavations on the Southern Parallel Feature, FAB e News 15 Autumn 2016
8. Excavations on the Northern Parallel Feature, FAB e News 15 Autumn 2016
9. Richard Whaley, Chawton Park Wood Air Photo Map, NEHHAS FAB e News 4 Spring 2012
10. Richard Whaley, Southern Parallel feature in Chawton Park Wood , NEHHAS FAB e News 10 Winter 2014-15
11. A31 East out of Winchester, Roman Road Abstracts, NEHHAS FAB 2016