Headington’s Great Lost Roman Pottery

Headington’s Great Roman Lye Pottery

www.headingtonheritage.org.uk/romanpottery Oxford Headington Roman Pottery at the Churchill Hospital Parchment Ware Headington headingtonheritage.wordpress.com/romanpottery Roman Kiln
Churchill Hospital Staff Car Park A – Once a great Roman Pottery


Perched high above the Lye Valley, sticking out like a tongue at the Lye Valley near the confluence of Boundary Brook, Churchill Hospital Staff Car Park A is an unprepossessing, ugly expanse of charmless aggregate, a depressing monument to environmental vandalism – damaging our planet, our children’s health, creating damaging stormwater runoff and crushing the life out of the calcareous springs feeding the ice age relict fen below. It is a place to stand and shudder.

Hard it is therefore, to imagine at this very place the hustle and bustle of one of the greatest potteries in Britain in Roman times, potters toiling with clay along tracks, hot kilns with smoking vents, heavy smoke rolling down the valley, the hum of the potters’ wheels, the curses from burnt hands, heavy loads, mud, ditches, and hard labour in freezing cold and heat – to create pottery sold far and wide around then Roman province of Britannia.

It is perhaps fitting to record this as we approach the 50th anniversary of the excavations led by Dr Christopher Young, who, during a three season archeological rescue excavation in 1971-73 uncovered one of the most important, and complete, potteries in Roman Britain, primarily documented in three interim reports (Young 1972a/73a/74a) and his PhD (Young 1977 & 2000) and subsequently by (Durham 1984) and (Evans 2005)

Headington Heritage Web: www.headingtonheritage.org.uk Web: headingtonheritage.wordpress.com Email : headingheritage@outlook.com Twitter :@headingheritage
How The Oxford Mail reported the dig in 1971

The source material, due to its highly detailed and technical nature is not easily accessible to a non-archaeologist.

The only contribution this wholly unqualified author can make is to introduce this important, but largely forgotten, story to a broader audience using today’s technology, thus, rendering it more digestible for the lay reader, a witness to the light but not of it.

The story of the Headington’s great Roman Lye Pottery at the Churchill Hospital, is one the earliest, and most important chapters in Headington’s long and proud history, deserving more prominence than it has hitherto received so it is hoped this article will address this injustice.

This is the first, not the last, published version of this blog which will expand as more material becomes available after COVID-19.

General Help

It can be useful to open maps and images in another window for reference by right clicking and select “open in new window” depending on your browser.

A glossary of the specialist vocabulary is provided at the end.

The Plight Of the Lye Fen

The Lye Pottery is directly over the Lye Valley SSSI, a precious, and very rare, ice-age relict fen under intense development pressure, and at immediate risk of destruction by Oxford City Council which is placing it in great danger by ill-thought out and reckless development.

Please see Warren Crescent 10 House Development – Risky, Reckless and Ruthless and The Hymn Of The Lye.

Please Join Friends Of Lye Valley (FOLV) to help save this precious place from Oxford City Council and other developers, see articles under menu “Lye Valley” on this site or email: friendsoflyevalley@yahoo.co.uk

Setting The Scene

Julius Caesar invaded, and promptly left, Britain, in 55 and 54 BC, but the main invasion, started in 43 AD was largely complete by 87 AD – it is clear that from at least Caesar’s time Rome exerted considerable influence on Britain’s Celtic tribes, both politically, but also bringing new techniques and industries, heralding the end of Iron Age Britain. By 410 AD the last legions had been withdrawn.

The Roman Landscape

This map shows known locations of Roman artifacts, mostly from the Oxford City Council area records only. (Radford 2012, Appendix 2) and basic features such as rivers and Roman roads.

The Roman Landscape – Google Map

Click top right on map to open it in new window or right click Oxford – Roman Landscape – Headington’s Great Roman Pottery or paste it into a new browser window.

Select features (eg: RK_9a) from the Google Map left menu (click top left to see if required) and zoom in to see specific details, click on the feature markers to see the details of each Roman feature.

References in this article, eg: (Map RK_9) refer to this map.

The Roman Landscape – An Overview

Several great pottery centres existed – Colchester, the Lower Nene Valley (Northants.) Verulamium (Herts.) Hartshill/Mancetter (Warwickshire), Wilderspool (Cheshire), The New Forest (Hants), Norton and Crambeck (Yorks) (Swan 1984, 91)

From the mid-third to the end of the fourth century AD, Oxfordshire was home to one of the three or four major pottery industries of later Roman Britain, with a history of production back to the first century. Oxfordshire wares were traded across a broad band of southern and central Britain from the far south-west to Kent and East Anglia. Outlying finds reached northern Britain as far north as the Mull of Kintyre and even to the other side of the English Channel. (Young 2020, 89)

The Roman road built between Dorchester on Thames and Alchester (Map Dorchester-Alchester) ran along approximately along the line of the eastern bypass (A4142), opened up the area economically, creating new markets for pottery and other goods.

Alchester (Chester=Castrum or fort) was the greater, and nearer (6 miles), of the two with Abingdon, an unfortified settlement, nearby. The site of modern central Oxford was little more than marsh and scattered farmland and small settlements on the second gravel terrace as rising water levels lead to the abandonment of the flood plain. (Radford 2012, 6)

OVER the last half century, it has become clear that the Roman road from Alchester to Dorchester-on-Thames was lined for much of length by a considerable pottery industry and that most of this development was centred on the high ground now covered by the eastern suburbs of modern Oxford, on the gravel terraces just north of Dorchester and on the Berkshire bank of the Thames. (Young 1972b, 209)

Along this road, within 2-3 miles on either side, in particular around modern eastern Oxford, a pottery industry developed, such as the Lye Pottery and other notable sites at Cowley (St Lukes), Rose Hill, Nuffield Centre (ex Wingfield Hospital) and Littlemore. (See Map) The Pottery, with a total, of approximately 16 kilns (over time), was clearly a large concern, others however may well have existed where evidence may have been destroyed or not yet located.

From the evidence here collected it is clear that this district was the centre of a considerable potting industry. It is close to the outcrop of the Kimmeridge and Oxford clays, which are to be found, the one at Shotover Hill on the east, the other towards Oxford on the west (Manning 1898, 19)

In the immediate area of the Churchill, Percy Manning (Manning 1898, 19) recorded kilns at Harry’s Bears Pit (Map RK_2a) as follows:

Running in a southerly direction from this quarry is a shallow valley, [The Lye Valley] down which goes a footpath to Cowley Marsh. [via Boundary Brook] At three other places along this valley Roman remains are found. [Map RK_VS_1, RK_2b] On the western side of the valley, between the footpath and Warren Cottages, [(Map RK_Warren)] the ground is covered with fragments of Roman pottery, [Near or at Lye Pottery] chiefly the ordinary buff-coloured mortaria and coarse grey ware, There were, however, two or three pieces of that rare fine smooth black ware which resembles Wedgwood’s ” basalt ” ware

There were also many pieces of burnt or half-burnt clay bearing finger and tool marks, just like the pieces of lining from the kilns described above, probably shewing [sic] the presence of kilns here too.  (Map RK_21)

At the Nuffield Centre (formerly Wingfield) two possible kiln sites were found (OHER 3670) (Map RK_8a/8b). All the above indicate the Lye Pottery was part of a local group of potteries in addition to being part of the larger “northern group” of potteries to the east of modern Oxford. (Young 1977, 12)

A Roman villa with a kiln was excavated in fields near Wick Farm (Map, RK1_Jewitt) (Jewitt, 1851) , (Wic=Saxon, market) and a possible settlement or market at Bayswater Road near Barton (Map RK_7) (Atkinson 1947,1948)

In Roman times, the Lye Valley extended from the current head of the valley near the modern SSSI entrance off The Slade to the north as a wide shallow valley under modern Girdlestone Road, then running to the west of and parallel with The Slade up to Old Road. (Map Lye_Valley_Outline, Yellow)

Then an extant fen, it was fed by very slow moving surface streams and highly calcareous groundwater seeping out of the valley sides, it would have presented a substantial barrier to communication.

The Lye Pottery

www.headingtonheritage.org.uk/romanpottery Oxford Headington Roman Pottery at the Churchill Hospital  Headington headingtonheritage.wordpress.com/romanpottery Roman Kiln
The Churchill Hospital Roman Kiln Site (Red) looking North – Now Staff Car Park A

The Pottery location is striking, at the tip of a tongue of land sticking out from modern Old Road which to the west is bounded by Boundary Brook to the south and east, the Lye Valley, then fen with gently sloping sides. (See Taunt photograph )

In Roman times, the Lye Pottery site (red polygon is the excavated area) would have been gently sloping as is still visible today, towards the Lye and Boundary Brook. It is believed the Pottery did not extend west of Churchill Drive, although no evidence is presented. (Durham 1984)

The below Lye Pottery Site Plan shows the layout of the Pottery with colours referring to phases of activity (see below), where Early Roman=Yellow, Phase 1=Green, Phase 2=Red, and Black=Unknown.

Archaeologists refer to features with notation such as F.307. (F=Feature) This is modified to eg: (Plan K-307) or just K-307, where eg: K-,Kiln, W-,Workshop, D-, Dryer, P-, Platform. Numbers with N with eg: R-N01 are the author’s labels on the Plan.

Excavation Sites I-VIII (or Areas) were excavated by Dr Young as above with Sites IX, IXa, X to the west, excavated subsequently. (Durham 1984, Evans 2005)

The following plan and chronology below give a summary of the information which is derived from Young (Young 1972a,73a,74a,77) and (Radford 2012, 53, Fig 3), (Durham 1984) and (Evans 2005, Fig 2) and overlaid on the site as it is today using Google Earth (Image of 07/05/2017 US)

www.headingtonheritage.org.uk/romanpottery Oxford Headington Roman Pottery at the Churchill Hospital Headington headingtonheritage.wordpress.com/romanpottery Roman Kiln
The Plan – Roman Lye Pottery, Churchill Hospital, Headington

Parts of the Pottery to the south were heavily waterlogged as evinced by drainage ditches, very possibly in origin part of a Roman field system, and a probable footbridge. A NW-SE depression with a probable stream at its base ran through Site VII, marked as a blue outline with the stream line (Plan R-N01) as described by Young. (Young 1973a, 3) This ran via the deepening ditch (Plan R-N01) (Young 1973a F.204/F.222/F.712) to the boundary fence (parallel to SP80) as shown, as did a parallel ditch to its east. Footpath SP80 is, today very wet at these two spots (E-NO2) despite the car park crushing this important water source to the fen below. (Plan SP80)

Centuries of ploughing, the establishment of the Churchill Hospital in the Second World War, (Oxfordshire Health Archive 2020) intensive urbanisation, made ground on the site, which also forms the artificially steep north side of the Lye valley, Girdlestone Road, The Slade and Peat Moors estate, have rendered the landscape barely recognisable from Roman times.

The Wares

Pottery can be described by function, fabric (what it is made of), period and style, this is a highly specialised area, see (Young 1977) for an in-depth analysis. A very large number of pottery types were produced which is beyond the scope of this article. See (Young 1977)

White Ware and Other Mortaria

White ware is kitchenware such as mortars and bowls, using mainly iron-free white clay.

The mortaria below are all of the 3rd-century type 2 with a more or less stubby flange … and a small bead-rim which was merely pressed down onto the flange by the potter’s thumb to form a spout (Young 1973b, 109) (Also Young 1973c (Fig. 2, 8-11). produced after 300 AD (Phase 2)

The “standard” moratorium, produced in Phase 1, but common by Phase II (4b) was Type M22. (Young 1977, 76)

  • www.headingtonheritage.org.uk/romanpottery Oxford Headington Roman Pottery at the Churchill Hospital Headington headingtonheritage.wordpress.com/romanpottery Roman Kiln
  • www.headingtonheritage.org.uk/romanpottery Oxford Headington Roman Pottery at the Churchill Hospital Parchment Ware Headington headingtonheritage.wordpress.com/romanpottery
  • www.headingtonheritage.org.uk/romanpottery Oxford Headington Roman Pottery at the Churchill Hospital Parchment Ware Headington headingtonheritage.wordpress.com/romanpottery Roman Kiln
  • www.headingtonheritage.org.uk/romanpottery Oxford Headington Roman Pottery at the Churchill Hospital Parchment Ware Headington headingtonheritage.wordpress.com/romanpottery roman kiln

Parchment Ware

Parchment ware is an important sub type of White Ware, first named from pottery found in Oxfordshire, normally with red paint decoration, used for table ware with a smooth surface described as:

Bowls and jars in pale granular wares, often with darker painted decoration, produced in the Oxfordshire potteries (Oxon/GB) and distributed across southern England during the 3rd and 4th centuries AD. (Roman Study Group/potsherd.net)

  • www.headingtonheritage.org.uk/romanpottery Oxford Headington Roman Pottery at the Churchill Hospital Parchment Ware Headington headingtonheritage.wordpress.com/romanpottery
  • www.headingtonheritage.org.uk/romanpottery Oxford Headington Roman Pottery at the Churchill Hospital Parchment Ware Headington headingtonheritage.wordpress.com/romanpottery
  • www.headingtonheritage.org.uk/romanpottery Oxford Headington Roman Pottery at the Churchill Hospital Parchment Ware Headington headingtonheritage.wordpress.com/romanpottery

Churchill Wares

Red or orange “Churchill” wares are sandy and soft, orange to red with/out a grey core, frequently wide mouthed jars, beakers and flagons and flanged bowls, unique to the Oxford region (Young 1972a, 23) made of iron-rich clay.

The following gallery shows other wares produced at the Pottery:

  • www.headingtonheritage.org.uk/romanpottery Oxford Headington Roman Pottery at the Churchill Hospital Parchment Ware Headington headingtonheritage.wordpress.com/romanpottery
  • www.headingtonheritage.org.uk/romanpottery Oxford Headington Roman Pottery at the Churchill Hospital Parchment Ware Headington headingtonheritage.wordpress.com/romanpottery
  • www.headingtonheritage.org.uk/romanpottery Oxford Headington Roman Pottery at the Churchill Hospital Parchment Ware Headington headingtonheritage.wordpress.com/romanpottery
  • www.headingtonheritage.org.uk/romanpottery Oxford Headington Roman Pottery at the Churchill Hospital Parchment Ware Headington headingtonheritage.wordpress.com/romanpottery
  • www.headingtonheritage.org.uk/romanpottery Oxford Headington Roman Pottery at the Churchill Hospital Parchment Ware Headington headingtonheritage.wordpress.com/romanpottery

A Chronology

Early Roman – First Century (Yellow)

Very small, simple assumed kilns (Young 1974, 10) were found from the Early Roman period (“The minor period”) in the first century. (Plan K-804, K-827, Yellow) producing “Belgic” pottery, mainly jars, showing a Roman influence even at this early stage:

Gallo-Belgic pottery was one such import to appear in Britain from the last two decades of the 1st century BC and 1st century AD. Made in workshops spread across northern Gaul it represents the first mass-produced fine ware to be made in Northern Europe. The products are mainly tablewares: cups beakers and platters, often bearing the potters name stamp, usually in a black ware (terra nigra) or a red ware (terra rubra). These vessels represent a completely new repertoire of forms and a technology not previously known in Britain.  (Prehistoric Ceramics Research Group 2020)

A well documented empire wide recession was recorded between 180-240 AD (Young 1977, 235).

Phase 1 – 250AD -300AD (Green)

After approximately 250AD, the Oxford industry, “sprang suddenly into prominence” (Young 1972b, 211) with the Pottery producing white mortaria (kitchenware), and orange “Churchill” wares with a soft sandy fabric, orange to red. (Young 1972a, 21)

In the late first and second centuries, urban potteries seem to have been particularly important

The third century seems to have been a period of radical change and the pattern of production in the later Roman period is very different. Of the urban potteries only Mancetter and Durobrivae continued to export their wares widely. However, the pottery industry as a whole did not decline, but rather expanded as large rural centres of production developed. The most important of these was Oxford with its widely distributed colour coat and parchment wares. followed by the New Forest which dominated Wessex (Peacock 1982, quoted in Middleton 1987, 251)

The first phase, dated to the late 3rd century, produced mortaria and orange coarse wares (250-300AD) (Young 1972b, 211), most typically “a wide-mouthed carinated bowl, decorated with horizontal painted bands on the exterior and more complex motifs (crosses, swags, circles) on the interior.” (Potsherd)

White wares were produced throughout Phase 1 and Phase 2 at the Site. (Young 1977, 58)

Phase 2 – 300AD – 400AD (Red)

After a period of disuse, at some point in the late 4th century a pottery drier and kiln, producing mortaria, parchment wares (Young 1977, 80)] and coarse wares, were constructed.

Around the kiln was an extensive area of pottery tips. (Young 1972b, 211)

Chronology Summary

Early Roman(p2)

1st Century

Plan : Yellow, VIII

Two small simple kilns (Plan K-804, K-827) , turf,prefabricated furniture (Young 1974a, 10)

Belgic pottery similar to Dorchester

Phase 1(4A)

250-300AD (approx)

Plan: Green,I,II,III,IV

Four kilns (2 sets)

Stone bedded in clay chamber, tongue pedestal support oven floor, clay ovens Same design,diff. sizes (K-307) (Complex A)

Small clay kiln on site V (Young 1973a, 210) (Complex B) with workshop

K-703, updraft, perm. vent hole type, tongue pedestal

Platform – P-107/8 (Complex A)

Circ. building/workshop W-403 with Dryer (Plan D-414 for green pots) and well (Plan WELL-207) filled with Phase 2 near complete pots (Young 1972a, 16) Complex A)

Mortaria, hard, soft and sandy to touch white sometimes orange/pink on outside, orange on inside. White wares, orange jars and beakers. (Young 1974, 2)

Orange “Churchill ware”. Grey reduced ware, rouletted (Young 1973a, 2)

Phase 2(4B)

300-400AD (approx)


Clay kiln K-101-2

All clay kiln with flue, tongue pedestal oven floor with corbels (supports under floor from sides) permanent vent hole type (K-101/102) (Complex C)

Kiln (K-452) identical to K101 on site IV (Young 1973a, 212) (Complex C)

Kiln (K-N605) poss. Phase II, site VI ( right angles to others) Young 1973a, 213) after 300AD (Complex D)

Stone dryer (D-202), clay chest (Complex C)

3 dryers (D-609) in/near building/prep area (Map W-604,red dot) (Complex C)

Mortaria and parchment wares, and reducing wares. (Young 1972, 2)

Valens coin (368-374AD)

K-903,K-904,K-914 (Site IX) Durham, Complex D

K-910 4B Durham (Site IXa), Complex D

K-1001 4B Durham (Site X), Complex D

Young (Young 1977, 49) splits Phase 1 (Green/4A) into two workshop complexes (Complex A+B) and Phase 2 (4B/Red/Complex C+D), each consisting of a workshop, kilns, dryers, and other structures such as platforms, clay chests, wells, ditches and sumps each forming a coherent operational whole.

Preparation and firing areas were separate from each other. Complex areas are not shown on the Plan to reduce complexity.

NB: There is a confusing naming discrepancy between Young’s interim reports (Young 1972-74) and his PhD. (Young 1977) Early Roman=PhD Phase 2, and Phase 1=PhD Phase 4a and Phase 2=Phase 4b. See (Young 1977, 55). This article uses the interim report convention.

The Potters and their Practices

The Potters

It has long been known that the Oxford region potters stamped their wares at two periods. In the second century stamps were impressed upon mortaria and in the late Roman period upon red colour-coat vessels copying samianware. (Young 1973c, 228)

Potters stamps can mostly be described as either “nonsense”, quadpartiate (four parts) or illiterate, only a few are literate as below.

Vossullus Stamp (Cowley)- (Young 1973c, 229)

A residual moratorium stamped by VOSSULLUS//VASSILLUS (Young 1973a, 211 was found at the Churchill in a Phase II ditch (Plan S-V), and in addition, 11 more at St Luke’s in Cowley (Young 1973c, Green 1983b, 10) in a dump (F.20) from approximately 50 to 250 AD.

The rim-profiles used by Vossullus would best fit a date in the first half of the second century. (Hartley, See Appendix – Vossullus ) (Also Young email) However, the lack of second century pottery activity at the Lye Pottery indicates it is highly probable he did not work at Lye Pottery, and this is a stray. (Young email)

Vossullus appears to be a name of Celtic origin. The name was present in the Roman Province of Gallia Narbonensis. (Southern France) In Latin this could equally read UOISSULLUS, or JOISSULLS. (As in Ulius Caesar)

A potter’s stamp with PVCINI M (Pucini?) was found among a few pieces of pottery described as “Samian” ware (Manning 1898, 19) at Shotover – a type made in 100-180 AD.

A Graffito (scribble) interpreted as “Tamesubugus fecit” was found on a three separated potsherds of pottery type M21 made between 240-300 AD (Young 1977, 59-60 ,64) at the Pottery, as below:

Tamesubugus Graffito (D305955a) © Oxfordshire Museum Service @headingheritage

Tamesubugus is also of Celtic origin, but the name has strong associations with the local area, meaning Thames Dweller, or Striker/Warrior (Young 2020, 89). For a detailed discussion see (Finn 2013), however, the conclusion he is the earliest named citizen, appears questionable based on the evidence above.

Note: The earliest named resident of the wider area was a Roman soldier buried at Alchester called Lucius Valerius Geminus in the first years after the Roman conquest in AD43. (Young email)

Origins and Organisation

The presence of Vossullus stamps at both St Luke’s, Cowley and at the Pottery, is indicative of either a personal name, or possibly a group, implies a high degree of organisation and sophistication at least an area scale.

We can speculate this may have been a romanised Celtic carpetbagger on the make seizing a business opportunity in a relatively new Roman Province, possibly from the well-established St Albans potteries. (Hartley)

Further, it could suggest that different groups, or potters, shared the kiln as common resource, marking their pots for identification, or possibly working on a piece rate basis, rather than as a consolidated business. Possibly this changed after 170 AD.

It seems probable many potters were migrants bringing their skills (Frere cited in Young 1977, 232), with local help from such as Tamesubugus.

Young (Young 1977, 50) hypothesises a workforce of 10-12, producing very approximately 40,000 pieces per annum. This marks Roman potteries as one of the first truly mass production industries in history.

The vast amounts of pottery scatter or “wasters” illustrates the scale of activity, and possibly a substantial failure rate. Ironically this is our greatest source of knowledge, the Young excavations alone produced 250 boxes. A further 2899 pieces weighing in total 51.6kg were collected in 2005 alone. (Evans 2005, 17)

The potters must have been very highly skilled – taking natural clay, understanding how to process it though many stages below, then mixing it with grit, forming and firing the pots. At any stage in this complex and long process the product could be ruined, due to poor mixing, air bubbles, drying out, uneven heat, dropping or any other reason.

Small quantities of Nene, Samian and pottery from Cowley (Green 1983a) (in addition to the Vossullus stamp) was found at the Churchill, implying at least some contact with other sites.

It is difficult to see how such sudden growth [At Phase 1], including the introduction of completely new fabrics and type ranges, could occur on the sole initiative of the local potters. It is unlikely that they would have sufficient backing to carry the risk of such a venture into mass production or the ability to establish the marketing system necessary to achieve success. It seems far more likely that this sudden expansion, reflected by growth in the number of kiln site as well as the increased distribution, was the result of intervention from outside the industry by landowners or merchants who, realising the need for new sources of mass production of pottery, saw that the existing Oxford industry with its excellent communications and its abundant supply of raw materials was an ideal candidate for investment and expansion in the circumstances all that would have been needed would have been the establishment of a marketing organisation and perhaps some sort of guarantee to sell specific types of vessels if potters produced them. This would explain  the rapid development of new mortarium types and the production of the colour-coated and parchment ware fabrics. (Young 1972b, 211)

The cultural shock for the Britons must have been enormous – perhaps the closest analogy is giving a cartmaker a tour of a Henry Ford motor plant, or a 1970s secretary with a typewriter a modern word processor. Crucially, the technical gap was, as with the previous two examples, close enough to be bridged – they could, and indeed did, learn.

The Potters – The Skull in the Well

Roman Kilns Churchill Hospital Headington Heritage Web: www.headingtonheritage.org.uk Web: headingtonheritage.wordpress.com Email : headingheritage@outlook.com Twitter :@headingheritage ©OxfordMail
Alas poor Vossullus I knew him Well (Arggghh….)
Chris and Judy Young examine a severed skull at Churchill Hospital, Headington, 1971
©OxfordMail @HeadingHeritage

Decapitation was was an honourable way to kill a defeated enemy or a way of ensuring the living would not return to disturb the living, however the placement of a decapitated head in the Phase 1 well (Plan WELL-207) with the front part of a sheep can only be seen as a very hostile Godfather style act both to the victim and the Pottery as a whole, perhaps leading to it being blocked in Phase 2. (Young 1972a, 16)

Processes and Materials

The following sections describe the processes, materials and artifacts used in broad pottery manufacturing sequence.

The area surrounding the Pottery is geologically complex comprising of thin strata, exposed at different locations as the land slopes down from Shotover, Headington and in the Lye Valley. Some of raw materials may have come from small, more local sources.

Clay – Provenance

www.headingtonheritage.org.uk/romanpottery Oxford Headington Roman Pottery at the Churchill Hospital Parchment Ware Headington headingtonheritage.wordpress.com/romanpottery Roman Kiln
BGS Geological Map of Pottery Surroundings

Clay, in particular, white clay came from nearby Shotover where a band of Kimmeridge Clay (for red and grey wares) (Young 1977, 12) surrounded the summit and stretched down southward closer to the other kilns (dark brown, left of blue above) in the Littlemore, Rose Hill and Cowley areas. (Young 1977, 11 Fig 3) Substantial deposits of white and yellow clay at the surface are still evident in BGS Borehole data. (BGS 1936, SP50NE2) (BGS SP50NE2, SPNE297).

Broadly, two types of clay were available to the Oxford potters. One, found only on Shotover, was iron-free and, when fired, produced pottery which is white or buff. [used in mortaria] The remaining clay sources of the region were all iron-rich and so could be fired under oxidising or reducing conditions to give on the one hand wares that were red or orange, and on the other those that were grey or black

Thus the clay allows a threefold division into white, oxidised and reduced wares. Each of these can be further divided on the basis of tempering, relative fineness or coarseness of the ware, and other special treatment such colour-coating. (Young 1977, 50)

Clay – Preparation

The preparation of clay of clay was a long and involved process, in addition to extraction and transportation.

Firstly, Weathering involved spreading the clay on the ground or platform (Plan P-107/108) and turning to break down particles and increase plasticity, then Souring in a cool damp place.

Levigation, or the placement of a clay slurry in a series of ditches would allow impurities to either sink or float. Drying would follow.

Next, Puddling, (in a puddling pit) meaning the mixing with sand as Temper, reduced thermal expansion and consequent cracking. (Swan 1984, 44-). This phase was particularly important for slips as it created a fine slurry. The resultant mixture could also be used as a waterproof lining for ponds, or for kilns and other construction. The rough texture of much of the output would indicate a high sand to clay ratio at the Pottery. (Young 1977, 18).

it is thought that raw clay was diluted with water in these puddling-holes [at Cowley] and mixed to a thin cream; the heavier impurities, such as sand and small stones, would sink to the bottom, leaving a layer of clean clay to be scraped off for use after the excess water had been removed by drainage and evaporation (Atkinson 1947, 12)

Finally, Wedging, or the cutting, beating and reversing of the clay, expelled air and ensured an even fabric texture.

The pots and mortaria would be made by Throwing on a wheel, followed by a slow drying process indoors.


Sand was used as Temper, probably from local sources.


Grit in the context of Romano-British pottery industries was used to improve the grinding qualities of the mortaria and was sourced from Lower Greensand deposits (eg: Boars Hill) (Young 1977, 11 Fig 3) (idem, 12) well to the south of Oxford, alternatively, closer smaller sources may have been available. (See Puddling above)


Water was required considerable quantities in multiple phases of the process. Highly calcareous streams and issues emerge along a springline at the top of the Lye Valley due to from a perched water table with one of the strongest described above at the Pottery itself.


Wood was used both as fuel and a construction material, may have come from woodland on the surrounding Headington Moor, or possibly from the valleys around, coppiced to ensure an ongoing supply of suitable size.

Peat and Fern

The Pottery was on a peat moor:

“The Poor” (Map, RK_The_Poor), extended from The Slade in the east, along the Lye Brook (Cowley/Headington Parish boundary), and up the western bank of the Lye, very probably providing an abundant fuel source for both kindling (fern) for the many drying process including for wood itself and more importantly, peat:

One Plot of Land and ground numbered 3 containing five acres situate in the Peat Moor ….Which said Plot of Land or Ground numbered 3 [The Poor] is in the judgement of the said Commissioners a full and ample equivalent and satisfaction for certain indulgencies usually had by the said poor Inhabitants in cutting taking and carrying away Furze and Fern and digging and carrying away Peat for their own use from certain parts of the said commons or waste grounds by the said Act intended to be divided and allotted.” (Headington Enclosure Award, 1804)

As the reference makes clear, the peat moor was more extensive than The Poor itself, probably also including the area of the Peat Moors estate on the other side of the Lye valley.


Shotover could also supply pigment in particular red and yellow ochre, as it has also for medieval wall-paintings (Edwards 1988) and this article from this site.

Building Materials

Correlian rag limestone, located 200 metres the north (Young 1972a, 10) was used for the base of some of the kilns and the various structures, such as flues and stokeholes.

Clay was used nor only to make the pottery, but also the kilns (with/out stone courses) their superstructure, chests and for waterproof lining of ponds.

Pottery Dryers

www.headingtonheritage.org.uk/romanpottery Oxford Headington Roman Pottery at the Churchill Hospital  Headington headingtonheritage.wordpress.com/romanpottery Roman Kiln
Phase 2 Pottery Dryer (Annotated, Enhanced)

Dryers were used to dry green, or freshly made, pots prior to firing. Several types of dryers were found, mostly T shaped driers.

A Phase I Dryer (Plan D.414) was set into the ground as a trough within the Phase 1 workshop.

The Phase 2 Dryer (Plan D.202) illustrated above, was the the largest at 4m*2.42m. A fire burnt at the base of the T descender inside the dryer (Young 1977, 20) as shown by a red dot above, fed via a flue from the stokehole outside. This would draw warm air from the flue to the T crossarm to the vents at each end of the horizontal bar of the T, which spread and slowed the flow. The pots would sit on a platform or pot tray above (light brown), constructed of a flimsy material of which no trace remains.

Identical T Shaped corn dryers were also used for corn drying at this period, the presence of Roman field boundaries, and the gap between the Early Roman period and Phase 1 and 2 may suggest dual usage, particularly during the recession of 180-240 AD.

Setting and Firing

Setting, or loading the pots in the kiln would be done with the pots upside down to form a series of domes or heat traps ensuring even heat distribution during firing.

An open permanent vent hole would produce oxidised pottery, if sealed with blocked peat or turf during the firing, reduced ware.


Early Roman (Yellow)

Early kilns (Yellow) (Plan K-804, 827), were probably of prefabricated construction with turf superstructures. (Young 1974a, 10)

Kilns – Phase 1 (Green)

The images below show the same single chambered kiln (Plan K.307) one of a group four Phase 1 kilns. (Young 1972a, Fig 3):

The reconstruction of K.307 below is based on diagrams (Young 1972a) with a probable superstructure above the oven floor as per (Swan 1984, 30 Fig II iv)

Phase I – Kiln F.307 Reconstructed as per Swan
Phase I – Kiln F.307 – Colourised from (Young 1972a)

These four kilns of the late third century (Young email) (eg: F.307) consisted of two pairs, each consisting of one large and one small kilns operating in the same period for large and small pots respectively. As the first pair fell into disuse or collapsed, the second was used. (Young 1977, 49).

In 1953, (Oxoniensia 1952-3, 224) two kilns (K-1953-1 (estimated position), and K-305) were found, the first probably destroyed as part of the Blood Transfusion Building (Plan BTU) construction, and the latter rediscovered by the gardener, Mr Sperrin, and covered. (Oxoniensia 1955, 90) Finds are in the Ashmolean (Register 1955 497, 618c) (also 1961, 521)

K.305 was then rediscovered by (Young 1972a) by his observation of an irregularly planted apple tree in a row (Plan K.305). A kiln found in 1962 in this general area was destroyed (Young 1972a, 12)

Fuel was inserted from the stokehole, a shallow pit in front of the kiln, into the furnace chamber. A horizontal flue between these was frequently present but all but absent in this case.

Above the furnace chamber was the oven floor or tray, supported, in this case, by a clay tongue pedestal and ledges at the sides. The walls were stone revetted with clay. Pots were placed into the oven (potting chamber) made of clay and potsherds.

Kilns – Phase 2 (Red)

The kilns at the Churchill site from Phase 2 differed in that they were mainly constructed of clay, with support from limestone rafts and had a flue. eg: (Plan K-101) Unusually, tongue support for the oven floor was supplemented by corbels (jutting out below like cogs). All were all single chambered (Young 1972a, 19)

Phase II – Kiln K-615 (?)
Phase II – Kiln F.615

The five kilns in Sites IX (K-903,K-904,K-914), IXa (K-910) and X (K-1001) have been identified as belonging to Phase 2 (4b) Complex D (Durham 1984) and by Green (Durham 1985, 100).

Kiln K-910 was built in the stokehole of K-452 although both belonged to Phase II (4B) (Durham 1894, IXa)

The orientation of kilns for both phases is mostly NW-SE probably to be set level in the sloping ground, however three Phase 2 Kilns K.615, K.904 and K.1001 to the west are orientated NE-SW as noted by (Evans 1984, 5)

Very similar kilns were found at Cowley (Young 1973b, Green 1983) and Dorchester (Harden 1936) For further information general on Roman kilns see (Corder 1964) and (Swan 1984)

Workshops, Structures & Storage

Both phases had stone circular workshops, Phase 1 – (Plan W-403) and Phase 2 – (Plan W-N04) with wooden superstructures in preparation areas with dryers set in the floor surface, as above Young identified a total of four complexes – A,B,C and D. (Young 1977, 49)

The subsequent excavation (Durham 1984) of Phase 2 kilns at Sites IX, IXa and X with almost no other supporting features (workshops, dryers etc) adds complexity to the original site interpretation (Young 1977) due to their distance from any workshop complex.

Clay and stone chests, clay-lined tanks, a platform (7m*8m) used for preparation, drying, working and storage (Plan P-107/108) (Young 1972, 16) were found in addition to a well (Plan WELL-207) containing some pottery from Phase I but mostly from Phase 2. (idem)

Where, Why and What

Where – Location, Location, Location

The Pottery location near the Dorchester-Alchester road (Map Dorchester_Alchester Roman Road) , only six miles from Alchester, the presence of white clay and pigment at nearby Shotover and good access to water have already been noted, however Roman kilns have been found at all of the above locations, so why here?

It is of note that the Alchester-Dorchester road had a distinctive westward bow, with its apogee where Old Road crosses the A40 just below Shotover.

Extensive Roman pottery and other finds on the north-west slope of Shotover Hill are reported by Manning. (Manning 1898) (Map RK_Manning)

Whereas this bow could simply be a means of circumventing Shotover Hill to the immediate east, it may possibly have significance as a direct route was feasible and for which the Romans had a notorious predilection for.

Another factor may have been that the Oxfordshire area was marginal to the Iron Age tribal territories of the Dobunni (to the west), the Catuvellauni (to the east) and the Atrebates (to the south) (Young 2020, 10). Very possibly members of one tribe would be reluctant to transgress into the territory of another. Shotover Hill, a very notable feature in a very gently rolling landscape, could well be of significance in this context.

At Crambeck the Pottery industry seemingly straddled the boundary between the Iron Age tribes of the Parisi and the Brigantes (Evans 1989, 43), thus making Crambeck ware accessible to two tribal markets. (Wood 2016, 87)

The Lye Pottery could have been connected to the Alchester-Dorchester Road and Shotover by a track (Map Roman_Way) (light brown) (Ordnance Survey 1900) and below, that led from the Lye Pottery leading north-east, following the Lye valley sides (Map Upper_Lye, Yellow), turning sharply north, and parallel with The Slade (Slaed, Saxon – Wet Valley) to the modern Nuffield Centre via the west side of the garden of 82 Old High Street where the last surviving section is still visible (Map Roman_Way_Remains) to avoid the Lye Valley (Map Lye Valley_Outline, Yellow) and below.

The below also illustrates the route of the possible Roman way going along the small ridge between the Upper Lye valley (parallel to The Slade) and a small, now buried, valley to the immediate west. The purple section still remains. (Source: EAW050074 )

Possible Route from the Pottery to Shotover, Dorchester-Alchester Road https://www.britainfromabove.org.uk/en/image/EAW050074 ©English Heritage

There is currently no evidence (pending investigation) that this track existed in Roman times, although the route is the most logical one – although the resident of 82, Old Road found a large white clay object in the early 1970s in her garden and gave it to the Oxfordshire museum.

A small Roman branch road from the Alchester-Dorchester road in Open Magdalen Wood (51.748233,-1.1955955,42/55630581) may have served the potteries, but the Lye Valley would have proved a barrier for the Lye Pottery. (Linington, 1959)

A possible roadside settlement at Bayswater Hill, Barton (Map RK_7) and the Roman villa (with a kiln see above) to the north-west of Wick Farm (Map RK_1,/RK_Jewitt) and other settlements may have provided limited markets or outlets.

Access to the Thames, particularly in light of recent discoveries of pottery near the Cherwell-Thames confluence, (Young 2020) would have been also beneficial.

The confluence of the Thames and Cherwell, lying 2 miles to the west of the site, may have been the highest navigable point. (Young 1977, 7) Transport of goods via the river from here to Londinium (London) appears logical. Very recent finds of pottery at Fairacres Convent (Young 2020, 89) (Plan RK_Fairacres) give limited support to this – the surrounding area yielded much Roman pottery:

The pottery assemblage was domestic in content rather than characteristically kiln-related, but included mortaria of the types produced in the Oxfordshire industries, so the Roman settlement in Fairacres was likely to have been a farm connected in some way to the potteries. (Griffiths 2020, 63)

Insufficient evidence exists to determine if the direct route from the Lye Pottery to the Thames/Cherwell confluence via Boundary Brook would have been usable, due to the presence of the then extant Cowley Marsh then at its base.

Why -Raw Materials

Vast quantities of fuel, clay, water and other raw materials would be required all of which were abundant in the area. Peat, in particular would have conferred a huge advantage, enabling year round production in a very wet climate.

Wood is an eminently suitable fuel provided it is throughly dry. If it is even slightly damp, the heat required to evaporate the water will tend to cancel any gain in temperature – and important factor in heat generation. (Swan 1984, 7)

So it is possible that peat could be used either as a fuel in its own right for the many drying processes required, or as a means of drying wood to make it highly combustable.

Local water is of particular importance as very considerable quantities are required, and it is difficult to transport even over short distances.

What – The Products

The distinctive sparkle of the output must have been an attractive feature for any buyer:

White clay enabled the production of white mortaria which was “soft and sandy to the touch.. ..

The most distinctive feature of any Oxford mortarium is the gritting which can be red, pink, white, grey or black in colour. It is always translucent and it always sparkles if caught by the light” (Young 1972a, 23)

Local Grit from Boar’s Hill, is crystalline and colourless when first dug up, but develops colours on firing as above. (Young email) Clearly, this would be highly marketable as would be the smooth surface of Parchment Ware. (idem)

The highly unusual calcareous composition of the water may have have acted as a defloccant, and added to the sparking effect, and hence added to the marketability.

An Appeal

This site, in conjunction with the other Oxfordshire kiln sites was the fourth largest Roman kiln industry in Britain – the excavations of the Pottery remain one of the largest ever undertaken, forming one of the most complete pictures of a Romano-British pottery we have.

This story is not Over….

It is hoped this article will rekindle interest in this important story.

Much MUST to be done before we can turn the last page – much of the good work done is in great danger of being lost forever without urgent work to preserve what remains.

A final report was never produced, no publicly accessible complete plan of the site is available, rendering interpretation of the key reports problematic. Other than interim reports, very little has been digitised or collated and indexed.

There is, at time of writing, yet an opportunity to gather the material with the assistance of those who were involved while the material and knowledge is still available – In effect, rescue archeology is required to rescue the records that remain, while they remain.

The To Do List

Oxfordshire Museum Service holds 365 records from the Churchill Site (Lye Pottery) indexed only by physical descriptions (eg:box xxx, bone, “lots of pots”) but with no interpretative classification, mostly with 1971-73 descriptors.

Sarah Green (Green 1983a) processed, classified and digitised 250 boxes from the Young excavations (1971-1973). The digital catalogue has been lost but a printout, believed to be in Oxford Archaeology, may well remain which could be redigitised.

The catalogue and the contents of the boxes would then need to be matched and documented – many of these probably are the source for images in Oxoniensia and other publications which could be redigitised with fresh photographs.

Finds by (Evans 2005, Appendix 2), numbering in the thousands, appear to be uncatalogued or referenced at Oxford Museum Service, only three references to 1984 are present (Durham 1983)

Clearly there is a need to document and catalogue all artifacts held by the Oxfordshire Museum Service, (See Artifacts section below) and match them to sites and excavations. An entire kiln was apparently dismantled on site, and given to Oxford Museum Service and since dismantled or disposed of.

Catalogue and digitise all remaining records, in particular perishable items such as slides, site plans, notes etc – some held at Oxford Archaeology, some others the author hopes to make available here in the near future.

Kilns K-914 and K.1001 are apparently still buried in situ and could reexcavated at some point in the future.

Write up a form of final or summary report and make it and the artifacts digitally available, digitise the PhD (Young 2000)

The remaining section of the possible Roman road (Map RK_, Purple) on the western side of the garden 82, Old Road could be excavated to determine if it could be a Roman road or track.

The draft report of (Durham 1984) which was never finalised, has been digitised by this author from draft notes and has been submitted to Oxford Archaeology.

There are modern potters making Roman pottery from kilns constructed using Roman techniques, (eg Potted History) but there are very few references, particularly in Britain, to sourcing the naturally occuring raw materials processing them to make pottery – this would shed much light on the skills, processes and materials required.

It appears funding for any of the above would be hard to obtain. Perhaps other resources such as PhD students could pick up the baton and engage in further study, some funds would be required for digitisation and other activity. The author will engage as COVID-19 restrictions are lifted, time permitting.

The author is happy to report Dr Christopher Young is making every effort to address the above.


It makes the author’s blood boil that one of the most archeologically significant sites in Headington was sacrificed for some buildings that could have gone elsewhere and a car park. What is lost is lost, impossible visions of happy schoolchildren being led around a well preserved site puzzling over the intricacies of kiln design, pottery manufacture and T shaped dryers, still arise, but alas, this will never exist outside of the imagination.

We should not let the story of Pucini, Vossullus, Tamesburgus and the many other Roman potters, the earliest of Headington’s citizens who we can name, be forgotten.

Their work was backbreaking, dirty, hard labour where not a single muscle was spared – carting, digging, stocking, building, digging, throwing, puddling, turning pots, wading in ponds and ditches, burnt by fire and inhaling acrid smoke, yet highly skilled – their legacy is their work, our work is to gather up their traces and ensure they are never lost – tempus fugit


I am very grateful to all the many people and organisations who have assisted with the project, the extensive help given does not imply any endorsement of opinions expressed which are entirely my own, as are any errors.

I was very struck by the unfailing friendless and generous help given by so many people, so freely, the article has been transformed from humble beginnings. I would like to thank in particular the following:

Dr Young for providing very generous help with proof-reading, photos and answering endless emails, as a result, a large number of misinterpretations and omissions were rectified, and the scope widened to include the post 1973 excavations.

Oxford Archaeology for locating the remaining records and permitting usage of the colour photo and drawing of kiln F.615 and the work by Nicola Scott to locate holdings. (which cannot currently be inspected)

Stephanie Jenkins, in particular for her knowledge of the Enclosure Award evidence of “The Peat Moor” and locating the “Tamesbugus made this” article.

Kay Hartley for giving her appraisal of the Vossullus stamp (see Appendix – Vossullus)

Angie Bolton of Oxfordshire Museum Service for the inventory listings and staff images of pottery used. As requested, for clarity, I must state these are not professional images and are not reflective of the normal quality the museum would use for public purposes.

Philip Kenrick, for many pointers, corrections, and tips and the the Oxford Mail newspaper article.

Peter McKeague for providing me with a draft report (Durham 1984) (never published) of the excavations of Sites IX, IXa, and X, this has been now been digitised.

David Radford, Archaeologist, Oxford City Council, tips, and permission to use the plan of the Churchill Site (Radford 2012, Fig 3) used as a base for the Plan , and the index of Kiln Sites (Radford 2012, Appendix 2) used in the Roman Landscape Map. The site dedicated to Vivian Swan’s work https://romankilns.net provided some extra details (also available via the SGRP web site https://romanpotterystudy.org.uk)

Oxoniensia, for generous free access to its earlier editions it provides, and permission to reproduce photos and plans. The article would not have been possible without this access and is in marked contrast to some other organisations (eg: British Archaeology) that impose heavy charges for access to material, which renders them unusable.

Thanks to Current Archaeology for sourcing, photographing and sending an article by Dr Young although normally chargeable content as Christmas present.

BritainFromTheAir for permission to reproduce the image extract above (EAW050074) at the maximum whole image size (584px max side).

The Oxford Mail, for permitting reproduction of both the dig report article and The Skull in the Well images free of charge.

The Ashmolean for providing me with copies of the accession registers of the 1953,1955 kiln artifacts.



Headington Heritage, A personal blog

Saving Headington’s Heritage
Visit  : www.headingtonheritage.org.uk

Visit  : headingtonheritage.wordpress.com

Email: headingheritage@outlook.com

Follow me on Twitter: @headingheritage 

YouTube: https://www.youtube.com/channel/UCJGrcWOjCV1ARRwUBoQhK5Q

Version: 1.0 12/05/21


ANON., 1955. NOTES AND NEWS. Oxoniensia, 20, pp. 90.

ANON., 1952. NOTES AND NEWS. Oxoniensia, 17, pp. 224-225.

ATKINSON, R., 1948. Archaeological Notes and News. Oxoniensia, 13, pp. 67.

ATKINSON, R.J.C., 1941. A Romano-British Potters’ Field at Cowley, Oxon. Oxoniensia, 6, pp. 9.

ATKINSON, R. and MCKENZIE, A., 1946. Archaeological Notes and News. Oxoniensia, 11-12, pp. 163.

BECKLEY, R. and RADFORD, D., 2012Oxford Archaeological Resource Assessment 2011. Oxford: Oxford City Council.

CORDER, P., 1964. The structure of Romano-British pottery kilns. Council for British Archaeology Research Reports, .

DURHAM, B., 1985. Churchill Hospital – Brian Durham. South Midlands Archaeology, 15, pp. 100.

DURHAM B./MCKEAGUE, P., 1984. Excavations at the Churchill Hospital, Oxford DRAFT. Oxford: Oxford Archaeology. (See Note)

EDWARDS, J., 1988. The medieval wall-paintings formerly at St Andrew’s Church, Headington, Oxford. Archaeological Journal, 145, pp. 263-271.

EVANS, D., 2005. Proposed South East Car Park, Churchill Hospital, Programme of Archaeological Recording, CA Project 1943 CA Report CA 05091
Oxford: Oxford Archaeology.

FINN, P., 2013-last update, Tamesubugus made this [Homepage of Archeox: The East Oxford Archaeology & History Project], [Online]. Available: http://www.archeox.net/sites/www.archeox.net/files/reports/Tamesubugus%20for%20web.pdf [12/01, 2020].

GREEN, S., 1983a. Churchill Hospital. South Midlands Archaeology, 13, pp. 135-136.

GREEN, S., 1983b. The Roman Pottery-Manufacturing Site at Between Towns Road, Cowley, Oxford. Oxoniensia, (XLVIII), pp. 1-12.

HARDEN, D.B., 1936. Two Romano-British Potters’ Fields near Oxford. Oxoniensia, I, pp. 81.


MANNING, P., 1898. Notes on the Archaeology of Oxford and its Neighbourhood. Berkshire, Buckinghamshire, and Oxfordshire Archaeological Journal, 4(4), pp. 9-28.

MIDDLETON, P.S., 1987.  The pottery industries of Roman Britain : a review
Céramiques hellénistiques et romaines. Tome II, II, pp. 252-266.

ORDNANCE SURVEY, 1900-last update, NLS Maps – Oxfordshire XL.SW, Revised: 1897, Published: 1900  [Homepage of National Library of Scotland], [Online]. Available: https://maps.nls.uk/geo/explore/#zoom=15.899999999999999&lat=51.75385&lon=-1.19470&layers=6&b=1&marker=51.878274,-1.169896 [10 December, 2020].

OXFORDSHIRE HEALTH ARCHIVE, , Churchill Hospital – Brief Summary. Available: https://www.oxfordshirehealtharchives.nhs.uk/hospitals/churchill [11th December, 2020].

PREHISTORIC CERAMICS RESEARCH GROUP, , GALLO-BELGIC POTTERY IN BRITAIN. Available: https://www.pcrg.org.uk/Articles/Gallo.htm [11/12/, 2020].

SWAN, V., 1984. The Pottery Kilns of Roman Britain. 5 edn. London: Her Majesty’s Stationary Office.

WOOD, R.J., 2016. Late Romano-British Pottery Production in Context: the Crambeck Ware Industry and its Landscape Setting , University of York.

YOUNG, C.J., 2020. The Oxfordshire Roman pottery industry. In: D. GRIFFITHS, ed, Archaeology of East Oxford Archeox: The Development of a Community. https://doi.org/10.5284/1081257 edn. Oxford: Archaeology Data Service (Dist), pp. 89.

YOUNG, C.J., 1986. The Upper Thames Valley In the Roman Period. In: G. BRIGGS, J. COOK and T. ROWLEY, eds, The Archaeology of the Oxford Region. 1 edn. Oxford: Oxford University Dept for External Studies, pp. 58-63.

YOUNG, C.J., 1977. The Roman Pottery Industry of the Oxford Region. 43 edn. Oxford: British Archeological Reports (BAR).

YOUNG, C.J., 1974. Excavations at the Churchill Hospital. Interim Reports III. Oxoniensia, 39, pp. 1-11.

YOUNG, C.J., 1973a. Excavations at the Churchill Hospital. Interim Reports II. Oxoniensia, 38, pp. 207-214.

YOUNG, C.J., 1973b. The Pottery Industry of the Oxford Region. In: A. DETSICAS, ed, Current Research in the Romano-British Coarse Pottery Research Report 10. St Andrews Place, London: Council for British Archaeology, pp. 105-115.

YOUNG, C.J., 1973c. The Roman Kiln Site at St. Luke’s Road, Cowley, Oxford. Oxoniensia, 38, pp. 215.

YOUNG, C.J., 1972a. Excavations at the Churchill Hospital. Interim Reports I. Oxoniensia, , pp. 10-31.

YOUNG, C.J., 1972b. The Oxford Potteries. Current Archaeology, 13, pp. 209-211.

Reference Notes

Dr Young’s 1977 PhD was reprinted in 2000 with a new introduction by the author and a bibliography of items published since 1976 by Paul Booth, this article used the 1977 edition only.

DURHAM, B/MCKEAGUE, P. – Excavation led by Brian Durham, 1984, draft report by Peter McKeague.



Kiln floor fragments, sherds from 1953 Accession No – 618c-621c, 1955 467 (Sperrin Kiln), 1961, 521.

The Lost Boxes of Finds and Records

OXFORD: Churchill Hospital – Sarah Green (Green 1983)

Processing and recording of the approximately 250 boxes of Roman pottery recovered from the excavations directed by Tom Hassall and Chris Young in 1971-1973 is now virtually complete. These records have been computerised using the Oxford University Computing Service computers – the ICL2988 and the Digital Vax 11/780; a preliminary catalogue of material sorted by context has been compiled. Initial data verification has been done using specially written SPITBOL programs and the data will be maintained and analysed using the SIR database management packages (PSTAT and GHOST 800) has been carried out. 

The only extraneous material that can be identified with any certainty are small quantities of samian, Black Burnished I amphora and Nene Valley pottery. Early material in the form of even smaller amounts of Middle Iron Age pottery has been noted by George Lambrick. Pottery occurs from other kiln sites within the Oxford area, for example, a 2nd century mortarium of type M2 with an illiterate stamp, possibly from Cowley. 

It is hoped that a programme of this sectioning at Southampton University will provide definite fabric descriptions and go a little way towards lightening some grey areas in Roman pottery studies – for instance in providing some help in distinguishing between reduced and oxidised wares made in Oxfordshire, Gloucestershire and Wiltshire. 

Two discoveries from the site which it is hoped will be followed up are: firstly a number of fragments of clay discs of between 20 and 30 cms in diameter and 1-2 cms thick, similar objects to which have been found at Farmoor (G Lambrick and M Robinson, Iron Age and Roman Riverside Settlement at Farmoor, Oxfordshire, CBA Research Report 32, 1979, p.54, fig 28), Pink Hill (Flood Plain Survey Site 179 see Lambrick forthcoming) and Tiddington, Warks (pers comm P Booth); one suggested use of which is that they are pot  136 lids. Secondly rims of large globular storage jars of distinctive form and fabric have been recorded, similar types having been found at Warborough, Cirencester (pers comm J Richardson), Rough Ground Farm, Lechlade, Alcester and Tiddington (pers comm P Booth) and their presence is suspected at Towcester and Dorchester. The origin of this distinctive type is so far unknown but one hypothesis as to its function is that it represents a sort of British Dressel 20 (an amphora imported to Britain during the 1st and 2nd centuries AD) having a very similar form and beginning chronologically where the imported amphora ends.

Oxford County Council – Museum Service

Records show Young, Durham and Evans deposited to the Oxford Museum Service. Much appears to be misclassified, and descriptions are only functional.

365 entries in Excel spreadsheet at grid ref SP 5465 0579 sites OX CCH etc using below site numbers, possibly this is the source material that Sarah Green computerised as above.

Churchill Hospital – Site Codes
Churchill Hospital
OX CHH 71 & 72
OX CHH 71 Tr I
OX CHH 71A & 72
OX CHH 71A & OX CHH 72
OXCHH 71 & 72 & 73
Churchill Hospital – Site Codes

Oxford Archaeology

Three boxes of slides, notes, and plans, possibly more. Colour picture of K.615 above. Very probably, a print out of the categorisation done by Sarah Green which appears to have been lost.

In addition, a now digitised copy of the draft report (Durham 1984) from notes and images. (Peter McKeague) the author will submit this to OU.


Coppicing is a traditional method of woodland management which exploits the capacity of many species of trees to put out new shoots from their stump or roots if cut down. In a coppiced wood, which is called a copse, young tree stems are repeatedly cut down to near ground level, resulting in a stool providing fuel.

Carinated A bowl with a sharp angle between the cylindrical upper part and the curving lower part of the body.

Corbel refers to an architectural member that projects out from a wall and acts as a type of bracket to carry weight, such as that imposed by a balcony above.

Fabric The material of which pottery is composed

OHER Oxfordshire Historic Environment Record.

Mortarium (pl. “mortaria”) was one of a class of Ancient Roman pottery kitchen vessels. They are “hemispherical or conical bowls, commonly with heavy flanges“, and with coarse sand or grit embedded into the internal surface. They were used for pounding or mixing foods and are an important indicator of the spread of Romanized food preparation methods. Stamps on some early Roman mortaria record the name of the potter, from which it is possible to trace their movement between workshops. 

Puddling Clay To puddle simply means to pound clay and water together in a dense mass which resists water penetration. Puddling breaks down the structure of the 
clay, closing fissures and forcing out air bubbles. The clay becomes very 
plastic, just as happens when it is made into pottery.  Dew ponds and other small water bodies  were puddled by sheep or human feet. 

Appendix – Vossullus

The following is an email from Kay Hartley, regarding the name Vossullus received by the author:


The second letter of this potter’s stamps is never as clear as one would like it to be, but I have always accepted VOSSVLLVS as his name, following R. J. C. Atkinson who was the first to excavate and publish a kiln and also pits, dumps and puddling-holes associated with his production on the site adjoining ‘Between Towns Road’ and St Luke’s Road at Cowley, Oxford (Atkinson 1941, 9-21). 

Atkinson had the clearest stamp, which I have seen (ibid., Fig 5, no. 44), and it is possible to see the west and east sides of the letter O, with the curves not completely joined at top and bottom. I agree with Atkinson that O is the best reading, but obviously if a much clearer stamp were found which differed, one would have to think again. The name is one of a number of variants related to a Celtic root. Atkinson looked the name up in Holder, Alt-Celtischer Sprachshatz who says that VOSSVLLVS and VOSSILLVS, are diminutives of a Celtic root VOSS–. The name VASSILLVS does, however, exist (Barnabas Lorincz 2002, Onomasticon Provinciarum Europae Latinarum Vol. IV, p. 184) mentions Vossillus in  Gallia Narbonensis, and Vassillus (ibid. p. 149, from Gallia Narbonensis, but also from BEG, (probably Belgica).

I have not seen any of his stamps from Sarah Green’s excavation (Green 1983), who followed Young 1973 in suggesting the possibility of alternative interpretation, Vassullus (Young 1973, p.228, no.2). The only example of his work known to me, from sites other than his production site is from Bishop’s Court Rectangle, near Dorchester-on-Thames (Oxon 42, fig. 9 and p.68).

The rim-profiles used by Vossullus would best fit a date in the first half of the second century.

References for above;

Atkinson, R J C, 1941, Oxon VI, 9-21

Green, Sarah 1983, Oxon XLVIII, p1-12

Holder, Alfred 1896, Alt-Celtischer Sprachshatz Leipzig

May, Jeffrey, 1977, ‘Romano-British and Saxon Sites near Dorchester-on-Thames, Ozfordshire’ in Oxon XLII, p42-79.

András Mócsy; Ortolf Harl; Barnabás Lőrincz 2002,  Onomasticon provinciarum Europae Latinarum (OPEL). Vol. IV :   Quadratia – Zures     Budapest : Archaeolingua Alapítvány, 2002.

Young, C J 1973 (pub 1974), Oxoniensia XXXVIII, 215-232


This text is licensed under this Creative Commons licence which permits, in summary, non commercial copying and reuse with attribution. The copyholders of images are marked eg: ©Oxoniensia and may not be reused without their consent, please email author for information.


Roman Kilns Churchill Pottery, Lye Pottery, Headington Heritage, Oxford Oxfordshire Roman Pottery Roman Dryers Headington Heritage