The Excavation of a "Red Hill" on Canvey Island

By Warwick J. Rodwell, F.S.A.(Scot.) dated 1966

Transcribed by Janet Penn

The "Red Hills" — General Introduction

Around the coast of Essex are an estimated two hundred "Red Hills": these are low, flat-topped mounds, which are very variable in shape and size and are situated on or near saltings. In extent they vary from less than a quarter of an acre to many acres and stand two to five feet high above the surrounding saltings. The body of a "Red Hill" consists almost entirely of loosely packed burnt earth of an orange or red colour. Mixed with this are many pieces of large, crude pottery vessels and fragments of clay bars of similar material. There are four different types of bars: "wedges", "firebars", "pedestals" and "T-pieces". All these, together with the pottery vessels, are collectively known as "briquetage." Sherds of Iron Age 'C and early Romano-British pottery, together with lumps of green vitreous slag, are also found in the burnt earth.

With a view to ascertaining the nature and purpose of these mounds, ex­tensive excavations were begun in 1906 by the newly-formed "Red Hills Ex­ploration Committee." Work was undertaken for four seasons on sites at Langenhoe, Goldhanger and Canewdon; two reports were issued. A summary of the "Red Hills" is to be found in the Victoria County History of Essex.

The widely accepted view on the purpose of the "Red Hills" is that they were industrial sites, producing salt from sea water by a process of evaporation. The crude pottery required for this purpose was produced on the site; slight traces of possible kilns and furnaces have been observed.

Photo:Fig 1: Simplified map of Canvey Island, showing important Roman sites

Fig 1: Simplified map of Canvey Island, showing important Roman sites

Warwick J. Rodwell

The "Red Hills" on Canvey Island

Canvey is a low-lying island near the mouth of the River Thames. Before the erection of the Dutch sea wall in the 17th Century the whole island was a salting and subject to periodic flooding. It has been established that the land level has sunk at least twelve feet in relation to the sea level since early Roman days. Our own calculations, based on excavation results, have verified this. The island seems to have been well populated from about the end of the 1st Century B.C. to the late 4th Century A.D. Finds of pottery have been numerous, par­ticularly on the "Red Hills." Fig. 1 is a simplified map of the island, showing the sites of the more important Roman finds. There are a number of "Red Hills" on the island and several more outside the sea wall, which have been partly or wholly destroyed by the sea. Several of these mounds were examined by Mr. Ernest Linder in the 1930s.

In the spring of 1964 the Wickford & District Archaeological Society began a long-term project of investigating the "Red Hills" in south-east Essex, particularly those which are being destroyed by modern development. During the first season excavations were undertaken on two of the Canvey sites: on "Red Hill" XII, where development is obliterating the mound; and on site III which is being rapidly destroyed by the action of the sea. The excavation on XII forms the subject of this report. All finds from the excavations are at present lodged with the Wickford & District Archaeological Society, but will eventually be deposited at the Chelmsford & Essex Museum.

The Excavation Site — "Red Hill" XII

Photo:Fig 2: Plan of "Red Hill" XII, showing excavation trenches in black

Fig 2: Plan of "Red Hill" XII, showing excavation trenches in black

Warwick J. Rodwell

This is the most westerly of the known "Hills" on Canvey: grid reference TQ 788820. It shows as a flat-topped mound, rising only two feet above the surrounding ground level, which is six feet above O.D. It is a particularly small site, being only one-sixth of an acre in extent; Fig. 2 shows a plan of the site. The main body of the hill is rectangular, but there is a considerable projection, at a slightly lower level, on the south-east. Upon this projection are two mounds, three feet high, which form the bases of W.D. huts. Both mounds were formed in Medieval times and have no connection with the underlying R.B. material. The site is divided by a wire fence running north-south and the western side has been planted with young trees. The west edge of the "Hill" was cut away during the digging of a massive dyke in 1963. This land is owned by the North Thames Gas Board and is the site of the new Liquid Methane Terminal. The portion of the site to the east of the fence is owned by Canvey Island Urban District Coun­cil and a new sewage disposal works is being constructed there. The Wickford & District Archaeological Society undertook rescue excavations on the site in the hope of establishing more facts about the dating and function of the "Red Hills" on the Island.

The land surface is composed of heavy green-grey alluvium; this is approxi­mately six to seven feet thick. Beneath it is a layer of blue clay, known to be as much as seventy feet in thickness. In Belgic and early Roman times the top of the blue clay appears to have been the land surface.

Seven trenches were opened and all taken down to the blue clay.

Trench C2A

Photo:Left Fig 3: Section through "Red Hill" (in block form. Trench C2A west face). Right Fig 4: Plan of postholes in north end of trench C2A, layer 4: taken on plane X-X.

Left Fig 3: Section through "Red Hill" (in block form. Trench C2A west face). Right Fig 4: Plan of postholes in north end of trench C2A, layer 4: taken on plane X-X.

Warwick J. Rodwell

This trench was cut through the small mound on which hut 1 was situated. The layers were of uniform thickness and a block diagram is shown in fig. 3. The first three feet were composed of alluvium, and contained over 2,000 sherds of Medieval pottery, ranging in date from 12th to 15th Century. Most of the pottery prior to the late 13th Century was crude, unglazed and of the cooking­ pot variety; after this date the wares were much finer, and often well glazed, the majority of sherds being from jugs. Animal bones, charcoal and shells of edible fish turned up in great abundance. Layer 3, a floor of burnt clay and charcoal, contained no finds, but as there were a few sherds of early Medieval pottery beneath it, then it too must be dated to that period.

Layer 4, alluvium, contained 1st and 2nd Century sherds at the bottom, and 3rd to 4th Century sherds at the top. A bronze coin from this layer was a Roman "as" of the 1st Century A.D. There was a complex of postholes and a gravel floor in this layer, see fig. 4. Postholes nos. 1 and 2 were larger and deeper than the rest, and both were at an angle of about 30° to the vertical, sloping towards the north. They may have been part of a retaining wall along one edge of the gravel floor. Of the other fifteen postholes all except one were vertical. It appeared that the posts had become broken off at ground level and oversealed by alluvium. Gradually the stumps of the posts rotted away, each leaving a cavity of the exact form of the post. At the bottom of most of the holes were fragments of rotten wood and bark. Many theories have been advanced as to the nature and purpose of these structures but none is sufficiently conclusive to be worthy of inclusion with this report. The structures are probably of the 2nd Century A.D.; closer dating is not possible due to the absence of associated finds.

A block of reddish grey lava, similar to pumice, was found; it measured 2 ¾ in. x 2 ¼ in. x 1 in. thick and showed clear marks of having been used as a sharpening stone for metal knives or similar tools.

Layer 5 contained remains associated with the salt industry, and was com­posed of loosely packed red burnt earth. Mixed with this were many fragments of briquetage, Belgic and Roman pottery of the 1st Century A.D.

Layer 6 was heavy brown clay, containing Belgic pottery and a few specks of charcoal.

Trench C2B

This was excavated on the north edge of the hut mound. The thick Medieval deposit was again encountered. The R.B. layers were more complicated, there being five layers of burnt earth, interspersed with alluvium. An early R.B. kitchen midden, containing, many shells, was sealed between these layers.

Trench C2D

This trench was situated just to the north of the "Red Hill, to ascertain if there were any outlying remains.  It revealed the site of a large Medieval hearth, which had been partly destroyed soon after its use. The shape and size was not clear, but it had been formed by many pieces of stone, including Roman materials. Much pottery, charcoal and animal bones were found, enabling the hearth to be accurately dated to the 13th Century.
In the alluvium below the hearth a few fragments of R.B. pottery were but found, but there were no domestic or industrial layers present.

Trench C2E

In order to examine further the hut mound this trench was opened between C2A and C2B. A complete Medieval circular hearth, three feet in diameter, was encountered. It was constructed of pieces of stone, flints and one fragment of Roman tile. Bones and charcoal were abundant, but there was no associated pottery. A posthole, six inches in diameter, was located beside the hearth and appeared to be contemporary with it.

R.B. and Belgic remains were comparable with those in trench C2A.

Trench C2F

The flat ground to the north of the hut mound was the lowest part of the 'Red Hill", and here the Medieval layers were very thin compared with those in the previous trenches. The remains of a much damaged Medieval hearth were found near the ground surface. The associated pottery dated this to the 13th Century. It had been of similar construction to those in trenches C2D and C2E, but contained considerably more Roman tile, some of which still had mortar adhering.

Late Roman occupation levels were again found overlying the burnt earth beneath which was a Belgic kitchen midden, the earliest remains found on the site. The section on the north face of this trench is shown in Fig. 5.

Photo:Left Fig 5: Trench C2F: section on north face. Right Fig 6: Trench C2g: section on west face.

Left Fig 5: Trench C2F: section on north face. Right Fig 6: Trench C2g: section on west face.

Warwick J. Rodwell

Trench C2G

The purpose of this trench was to ascertain the exact southern limit of the main body of the "Red Hill." Medieval and late Roman remains were few; an unusual find was a sherd of blue-grey colour-coated ware which appeared to be a waster from a kiln. The southern edge of the burnt earth was reached and the briquetage found included two pieces of "wedges". Fig. 6 is a section drawing of the east face.

Trench C2H

The layers and finds were comparable with those in the previous trench, and the tapering deposit of burnt earth indicated the west edge of the "Red Hill."

General Finds

All trenches produced vast quantities of animal bones; sheep and hare were the most common, but ox and pig were also noted in the R.B layers

Shells too were abundant, being mostly oyster, but with some mussel, cockle, periwinkle and whelk.

Although no flint implements were found, many flakes were present in the lower levels.

A number of small unidentifiable objects, together with some nails, were recovered from the R.N layers.

Two small pieces of sheet lead, one of which was rolled into the form of a short tube, were found in the upper Roman levels.

Burial Site
In 1963 the Society heard that workmen constructing the dyke had dis­covered "urns, glass vessels and heaps of crumbling bones". These were all broken up and scattered by the mechanical excavators. A few sherds, however, were rescued and enabled an approximate date to be obtained for the burial deposit.

Conclusions and Dating

The earliest remains on the site are those which are sealed below the burnt earth, showing that the site was occupied before the "Red Hill" was begun. Such remains were found in trenches C2A and C2F. The latter was a kitchen midden containing many oyster shells, pottery and a few bones. The earliest occupation may have begun at the end of the 1st Century B.C. The "Red Hill" came into existence during the early years of the 1st Century A.D. This date is supported by the abundance of Belgic pottery found in certain of the burnt earth layers.

The date when the "Red Hill" ceased to produce salt is difficult to ascertain; the pottery in the burnt earth layers stops at the end of the Flavian period, with the possible exception of one piece: no. 6a. It is an everted rim of a latticed olla, Cam. f.278. At least a dozen more examples of this form were found on the site, all of which were above the burnt earth. No. 6a, however, was in the middle of the burnt earth in trench C2H and was in association with no. 14 (bowl, Cam. f.37). Unfortunately this form of bowl was in use throughout the Roman period, and is of no help in close dating. The olla occurs most frequently at Colchester between 100 and 140 A.D., but how soon it started is uncertain. Many authori­ties claim that the "Red Hills" ceased to function by about 80 A.D., but the pottery from this particular site indicates a slightly later date. In the absence of more direct evidence, a date of 80 to 100 A.D. can safely be assumed for the end of the salt production on this site.

The pottery from the burial site in the dyke is mostly late Flavian, with some possibly Trajanic. A date of c. 100 A.D. seems likely for this deposit.

The abundance of pottery shows that occupation of the site continued throughout the 2nd and 3rd Centuries, and into the 4th Century. The structures in trench C2A almost certainly belong to the 2nd Century. Although no trace of foundations has yet been discovered, it is certain that a masonry building with tiled roofs once existed in the area; in the late Roman levels many pieces of tiles and building stones, some with mortar adhering, were found. The ruins of this building must still have been apparent in the 13th Century, as the Medieval hearths were entirely made of re-used materials.

The site appears to have been unoccupied during the Saxon period and the earliest piece of Medieval pottery is of 12th Century date. Two of the hearth sites belong to the 13th Century, the other is also probably of the same date. Trench C2A produced a good selection of sherds of glazed jugs of 14th and 15th century date. A thin scatter of 16th and 17th Century sherds was also noted.

Photo:Photo taken of the Red Hill near the Gas Works. Is this Hut one site of Trenches C2A/B & E?

Photo taken of the Red Hill near the Gas Works. Is this Hut one site of Trenches C2A/B & E?

Norman Chisman

Photo:Photo taken of the Red Hill near the Gas Works. Is this Hut one site of Trenches C2A/B & E?

Photo taken of the Red Hill near the Gas Works. Is this Hut one site of Trenches C2A/B & E?

Norman Chisman

This page was added by Janet Penn on 18/05/2013.
Comments about this page

Jan, Thanks for publishing both these articles on archeological excavations on Canvey, after the first appeared I hoped there would be something about the Red Hills in the west and centre of the Island but this detailed report on a site on the Methane Terminal came as a surprise.I'd always thought that Thorney Bay and creek might have been significant in the siting of the Red Hill area that Linder excavated in Blackmore Ave/Leigh Rd as it might have provided a safe haven for waterborne transport. However these finds seem to imply that due to higher land-level(almost 12ft) the bay probably didn't exist particularly as an occupied site is shown on the other side of what is now the deepwater channel.Post these excavations I believe there was a discovery of another Romano-British burial at the junction of Thompson Ave and Thorney Bay Rd so I guess that by putting all this together we could imagine a very busy industrial settlement in Romano-British times stretching all along Canvey's southern shoreline from the Point almost as far as Holehaven. Fascinating!

By Graham Stevens
On 20/05/2013

You can see a map showing the "Red Hills" locations and Canvey coastline as it was at that time in Dowd's Cyclopaedia

By Janet Penn
On 20/05/2013

Thanks Jan this map is really useful. I'm pleased to see that Dowd presumes that an inlet existed in the shoreline where Thorney Bay is today, this lends weight to my original hypothesis.Now more questions arise: 1/ Given the lack of obstacles north of the shoreline(i.e creeks)was there a road to Chelmsford(Roman garrison town)? 2/ Was there a connection between the salt-works and the Roman villa in Dawes Heath which was discovered by aerial photography in the hot summer of 1976? 3/ The biggest question of all. What was the reason for the dramatic land-fall or rise in sea-level at the beginning of the first millenium? P.S. First time I've been on YOUR updated and upgraded site. Very impressive. Congratulations! I've had some fun looking at Phil Blanchard's cartoons, trying to recognise any of the characters.

By Graham Stevens
On 20/05/2013

Wasn't there a Roman road from Canvey found? I have more Archeaolgy stuff to go on and I will also copy some of Dowd's over at some point when I am not so busy. Glad you liked the site Graham

By Janet Penn
On 20/05/2013

If you're already a registered user of this site, please login using the form on the left-hand side of this page.