Report on a ceramic assemblage from Clausentum

Attention: Important Disclaimer

This paper is a recent modification of a mock report on a medium-sized ceramic assemblage. The original report was a marked assignment for the University of Southampton, during the author’s MA studies between 2008 and 2009. The aim of the original assignment was to prepare the graduate to produce basic archaeological reports on ceramic assemblages. The pottery that is discussed in this paper is real and was handed to the author by a UK commercial archaeology unit, which still reserves the publication rights for the site report and its excavated material.

To avoid any complications, the author decided not to upload his original assignment on the web due to certain copyright issues and out of professional courtesy to the archaeological unit, which originally handled the assemblage. The current paper presents the author’s own analysis and discussion of the material, including personal illustrations and photographs, which are still the author’s own intellectual property. The location of the archaeological site, where the material comes from, and other information related to its excavation, have been purposely altered.

The author clearly states that the publication of his work on his personal website does not consist of a commercial activity and does not aim towards economic profit. By contrast, the author wishes his work to be publicly available for free and to be shared openly among those interested in reading it, particularly students in the analysis of ceramics and lithics, who might be looking for examples of finds reports.

Introduction

This report discusses the ceramic assemblage from 41-43 Ioannis Anglicus Road, Southampton. The assemblage was produced after two trenched evaluations in 2001 and a full excavation in 2002. The site has been identified as part of the Romano-British settlement of Clausentum.

The report is divided in three sections. Section 1 presents some general information on the sites and the history of the excavations. Section 2 presents the ceramic evidence divided by wares groups. A detailed catalogue of the pottery has been separately submitted in the form of an Access Database and is not included in the present report. Section 3 discusses the chronology and distribution of the ceramic assemblage in relation to the broader site.

1. The site

The area under investigation includes two plots of land, located at 41 and 43 Ioannis Anglicus Road, Southampton. Both plots have been identified as parts of the broader Romano-British settlement of Clausentum. The first excavations in the area began in 1937-8 by Waterman and continued in 1951-4 by Cotton and Gathercole (1958). These excavations revealed the earliest phases of the settlement dated in the late Flavian period, and more specifically towards the end of the 1st century AD.

In 1998, trenched evaluations conducted by Archaeology South in the neighbouring plot of land at 45 Ioannis Anglicus Road, identified part of a Saxon cemetery. Radiocarbon dates suggested that the cemetery was used until the late 7th century AD. In 2001, a new trenched evaluation at 41 Ioannis Anglicus Road (AS41/2001) produced no Saxon evidence and no human skeletal remains, which might have associated with the Saxon cemetery. However, the evaluation produced significant quantities of Roman pottery, which was mainly found in a drainage ditch and a rubbish pit. More specifically, the evaluation included 6 trenches: trench 1 at the front area of the old house, which produced most of the Roman pottery, and trenches 2-6 at the side and back garden of the house. From these trenches only 2, 3 and 4 produced ceramic evidence (after personal communication with Archaeology South, April 2009).

A second evaluation at 43 Ioannis Anglicus Road (AS43/2001), also in 2001, produced Roman pottery from a ditch running North to South of the plot; however, no evidence of the Saxon cemetery were present. This evaluation included 5 trenches: trench 1, located at the front area of the old house, and trenches 2-5 at the back of the old house and the empty space between the old house and the garage. Pottery was only found in trench 5 (after personal communication with Archaeology South, April 2009).

The 2002 excavation of both plots (AS4143/2002) revealed a single feature, which was a drainage ditch running between the two houses. The majority of the pottery recovered from the ditch was Roman, while few sherds were post-medieval to modern (after personal communication with Archaeology South, April 2009).

2. The ceramic evidence

The site produced a total of 285 sherds weighing 8,362 grams. More specifically, the first evaluation (AS41/2001) produced 165 sherds weighing 4,279 grams, which formed nearly 50% of the total assemblage by sherd count or 88.2% by weight. The second evaluation (AS43/2001) produced 52 sherds weighing 1,115 grams and the final excavation of both plots (AS4143/2002) produced 68 sherds weighing 2,968 grams. The total assemblage can be divided into to eight ware groups: amphorae, Samian wares, mortaria, imported fine wares, imported coarse wares, Romano-British fine wares, Romano-British coarse wares and modern ceramics.

2.1 Amphorae

Dressel 20

The site produced nineteen sherds of Dressel 20 amphorae weighing 1,867 grams. Such hand-made amphorae were manufactured in the Guadalquivir valley of Southern Spain, in the Roman province of Baetica. Such vessels were primarily used to transport olive-oil across the Roman Empire (Tyers 1996, 87). Their fabric matches BAT AM of the National Roman Fabric Reference Collection (Tomber and Dore 1998, 84). The sherds are thick with coarse sandy fabrics, often not well wedged, with large elongated voids. The BAT AM fabric is pinkish buff (7.5YR 7/4, 10YR 8/3) and rarely pinkish white (5YR 8/3). The sherds’ exterior surfaces are buff to off-white, sometimes covered in thin wash. Their main fabric inclusions are medium to small-sized quartz grains in various colours (white, grey and often white with red iron ore edges), limestone and sparse mica. Their external feel is rough and the sherds are hard.

Gauloise 4

Gauloise 4 amphorae number forty-seven sherds weighing 1,273 grams. These used to be wine amphorae manufactured in various locations in Gaul and are nowadays described as ‘flat-based’ amphorae (Tyers 1996, 94-5), although their bases are not exactly flat. Their fabric matches GAL AM I of the National Roman Fabric Reference Collection (Tomber and Dore 1998, 93). They are wheel made and easily identified by their distinct shapes and ring bases (e.g. Figure 1, n.5), which allow then to stand ‘flat’. Their fabric is generally fine, pale buff (7.5YR 8/5) or very light orange (2.5YR 6/8), medium hard and the feel of the sherds is medium rough. The exterior surface of some sherds carries a thin wash; those with wiping and smoothing marks have a medium smooth feel. Their main fabric inclusions are small and well-rounded quartz grains, small red and black particles, clay pellets, large bits of limestone and sparse mica, all mixed together in a silty matrix.

Unknown Type A, possible imported amphorae

Two sherds (19 grams) from ditch fill 113 in trench 1 (Evaluation SA41/2001), have been identified as possible imported amphorae. Both sherds come from the same wheel-made vessel. Their feel is medium rough and their texture is medium hard. Their fabric is medium to coarse, light red (2.5YR 6/6) and its main inclusions are large angular white and red quartz grains, up to 5% grog and scarce voids from organic impurities that burnt out during firing.

2.2 Samian wares

The site produced thirty-two fragments of Samian pottery weighing 295 grams. All Samian wares, recovered from all excavation phases, are of Southern Gaulish origin and more specifically from the kilns of La Graufesenque. Their fabrics match LGF SA of the National Roman Fabric Reference Collection. One sherd comes from Montans and matches MON SA of the National Roman Fabric Reference Collection (Tomber and Dore 1998, 28-9). The sherds come from various small bowls, cups and dishes, either plain or stamp-decorated, mostly with flower and animal motifs. Their external colour is dark-red and glossy (10YR 4/8), often highly lustrous and always very smooth. Their fabric is red to pale brown with very fine silty texture. Their main inclusions are fine rounded limestone particles and sparse mica. The sherd identified as Montans Samian is tempered with large quantities of fine limestone and red iron ore particles.

Ditch fills 107 and 114 in trench 1 (evaluation AS41/2001) produced two sherds from a small bowl with stamped floral decoration. Ditch fill 116 in trench 1 produced three sherds (Plate 1, n.17-19; Figure 2, n.1) from a Dragendorff 35 bowl (Tyers 1996, 108-9) dating in the Flavian period (c. 70-100 AD). Ditch fill 118 in trench 1 produced nine sherds (Plate 2, n.14-22) from a single Dragendorff 30 drinking cup (Tyers 1996, 109). The cup carries stamped floral decoration, bird and deer motifs, which are framed by curved lines and two incised rim bands (Figure 2, n.2). The cup is also Flavian (c. 70-100 AD). Ditch fill 125 in trench 1 produced a small rim sherd from an undecorated bowl. Pit fill 216 in trench 2 produced a rim sherd from a Dragendorff 37 bowl (Tyers 1996, 108-9) with no decoration, which dates between c. AD 70-230. Based on other ceramic evidence, it appears more likely to be Flavian.

Ditch fill 505 in trench 5 (evaluation AS43/2001) produced ten sherds (Plate 3, n.37-46) from at least three different vessels, most of which decorated with stamped straight and spiral lines, flower motifs and heart-shaped leaves, all dating in the Flavian period. One sherd (Figure 2, n.8) comes from a Dragendorff 27 undecorated bowl (Tyers 1996, 110) and another sherd (Plate 3, n.39) comes from a Dragendorff 30 cup (Tyers 1996, 108). Ditch fill 506 in trench 5 produced the only Montans Samian fragment, probably deriving from an undecorated dish.

During the excavation of the site (AS4143/2002), ditch fill 15 produced an undecorated sherd from a bowl (Plate 4, n.24); ditch fill 23 produced two decorated sherds of Flavian date; and finally, ditch fill 26 produced two sherds from an open bowl, probably a Dragendorff 35-36, also Flavian.

2.3 Mortaria

The site produced eight sherds from at least two mortaria weighing 1,718 grams. All mortaria are locally made and more specifically in the New Forest region. Such vessels have distinct forms (Tyers 1996, 125) and carry flint-tempered trituration grits on their interior surfaces. Their fabric is hard and sandy, containing abundant well-sorted medium-sized quartz grains up to 20%, medium-sized angular grey flint and sometimes grog. The colour of the fabric is white or pinkish white (2.5YR 7/6). The sherds’ feel varies from rough to smooth, particularly in fragments with wiping marks on their exterior surfaces . Trituration grits consist of large angular quartz and crushed flint particles in grey, white and black colours. The fabric of the sherds matches NFO PA, which is the New Forest Parchment Ware of the National Roman Fabric Reference Collection (Tomber and Dore 1998, 142). The same fabric was used for the production of local coarse wares, mainly bowls, which are discussed further below.

Ditch fill 103 in trench 1 (evaluation AS41/2001) produced four wall sherds and a rim fragment, 23cm in diameter (Figure 1, n.1). Ditch fill 505 in trench 5 (evaluation AS43/2001) produced two sherds (Plate 23, n.1-2) and ditch fill 506 produced one small spouted rim sherd of unknown rim diameter, which is heavily worn. During the excavation of the site (AS4143/2002), ditch fill 23 produced a rim sherd, 32 cm in diameter (Figure 1, n.2). The above fragments match the New Forest mortaria typologies by Fulford (1975, 74-9) and Tyers (1996, 125), dating between AD 260-370.

2.4 Imported fine wares

Central Gaulish black-slipped wares

The site produced eight fragments of Central Gaulish black-slipped wares weighing 173 grams. The sherds come from medium-sized thin-walled vessels, which are almost exclusively bowls. Their fabric is very fine, light pink (2.5YR 6/6) with glossy, almost metallic, black slip. Their main inclusions are fine sub-angular quartz grains, fine limestone and red iron ore particles. Their matrix is heavily micaceous, dense and well-wedged. The sherds are hard with smooth external feel and their fracture is snappy. Such vessels were produced at small scale in Central Gaul by terra sigillata workshops and their production began sometime in the early 2nd century AD (Tyers 1996, 137). The current samples match fabric CNG BS of the National Roman Fabric Reference Collection (Tomber and Dore 1998, 50).

Ditch fill 116 in trench 1 (evaluation AS41/2001) produced four sherds, one of which is a base fragment from a bowl, 7cm in diameter (Figure 1, n.3). Ditch fill 118 in trench 1 produced four fragments from a flat-based bowl, 8 cm in base diameter. The bottom of this base fragments carries an incised letter ‘K’, possibly part of the producer’s signature or the workshop’s initials. Such potter signatures are likely to suggest that these vessels were fired in a communal kilns.

Unknown Type B, possible imported fine wares

Ditch terminal fills 303 and 305 in trench 3 (evaluation AS41/2001) produced two fragments (10 grams) of an unlisted fabric of unknown origin. Both sherds come from the same vessel but are too small to determine its original form. It is only possible to say that they come from a wheel-made fine ware. The fabric’s texture is fine and the fracture shows four colour zones, fading from orange to pale pink. The sherds are covered in bright orange slip with abundant golden mica, which sparkles on the exterior surface. The fabric’s matrix is silty with very fine small-sized grey and white quartz grains, iron ore, clay pellets and mica up to 20%. It is not entirely certain if this fabric is imported as there is some possibility to originate from areas with highly micaceous clays in Britain (e.g. Dorset or Cornwall). Again, the same geological formations are also common in the opposite European coastline and further inland.

2.5 Imported coarse wares

North Gaulish white wares

The site produced four sherds of North Gaulish imports weighing 42 grams. Such sherds come from small-sized table wares, possibly bowls or jars, which are poorly preserved. Their fabric is cream (10YR 8/3), medium-soft and their feel is difficult to determine due to surface deterioration. Their main inclusions are well-sorted sub-angular quartz grains, sparse limestone and very small red and black particles, probably iron-rich, up to 3%. Their fabric matches NOG WH 4 of the National Roman Fabric Reference Collection (Tomber & Dore, 1998, 75).

Ditch fill 505 in trench 5 (evaluation AS43/2001) produced two decorated sherds (Plate 3, n.10-11). The largest one carries distinct cross-hatched decoration forming rhomboid patterns (Figure 1, n.9). This specific motif was popular in both Gaulish and British decorated forms of the Roman period (Tyers 1996).

Unknown Gaulish grey coarse-wares

A single sherd weighing 9 grams was recovered from ditch fill 116 in trench 1 (evaluation AS41/2001). This fragment was identified as imported grey coarse ware, possibly originating from Gaul. The sherd is light grey (7.5YR 6/0), wheel-made, with no external treatment or decoration. Its fabric resembles the Trier black-slipped ware (Moselkeramik) (Tyers 1996, 138). It is very hard, fine-textured, with a distinct ‘sandwich’ core of dark red between grey margins. Its main inclusions are fine quartz grains and very small black particles, most likely ferrous, all evenly mixed together inside the matrix. The main difference between this sherd and the Trier black-slipped ware is that the former has no external slip; it has only been wiped hard with a piece of textile, which left distinct surface marks.

2.6 Romano-British fine wares

New Forest colour-coated wares

Ditch fill 103 in trench 1 (evaluation AS41/2001) produced two sherds weighing 25 grams, which were identified as New Forest (metallic) colour-coated wares. They probably come from the same vessel, 20-24cm in external diameter, and carry white painted decoration around the neck area. Fulford (1975) identifies three main shapes produced in this fabric: beakers, jugs and flasks. However, these sherds are also likely to come from an open vessel form. Their fabric is extremely hard and high-fired, and according to Tyers (1996, 171), such vessels can be described as stone wares. Their interior and exterior surfaces are dark grey with metallic sheen. Their fabric inclusions cannot be distinguished due to fusion. The fabric matches NFO CC of the National Roman Fabric Reference Collection (Tomber and Dore 1998, 141).

New Forest brown-slipped wares

The site produced six sherds (29 grams) of New Forest brown-slipped wares. All come from small-sized bowls or jars with 3mm average wall thickness. Their fabric is very fine, white and hard. All sherds carry dark reddish-brown slips. Their only visible inclusions are very fine small-sized quartz grains and their fabrics match NFO CC of the National Roman Fabric Reference Collection (Tomber and Dore 1998, 141).

Ditch fill 107 in trench 1 (evaluations SA41/2001) produced four sherds, which come from a single bowl, 8cm in rim diameter (Figure 2, n.8). The bowl belongs to Fulford’s Type 57.4 (Fulford 1975, 63) with its characteristic everted and pointy rim, dating in the Late Roman period. Another two sherds were recovered from ditch fills 15 and 23 during the excavation of the site (AS4143/2002) (e.g. Plate 4, n.23).

Unknown Type C, possible local fine ware

Ditch fill 505 in trench 5 (evaluations SA43/2001) produced three sherds (29 grams) of an unlisted fabric (Plate 3, n.4, 6-7). These include a flagon’s neck with no decoration and two wall shreds from a possible cup with relief dotted decoration. All sherds are wheel made, smoothed, with light thin orange wash. Their fabric is very fine-textured and light red (2.5YR 6/6 to 2.5YR 6/8). Its main inclusions are small to medium-sized sub-angular quartz grains, red iron ore, small black particles, sparse mica and limestone. It is not clear if these sherds are imported or local. Judging from their thin walls (3mm) and their dotted relief decoration, including a circular relief motif on sherd no.7 (Plate 3), it is more likely to associate with the New Forest industries.

Oxfordshire red-slipped wares

Pit fill 212 in trench 2 (evaluation AS41/2001) produced two sherds of Oxfordshire red-slipped wares weighing 10 grams. The sherds come from two different wheel-made inverted beakers. Their fabric is hard, orange buff with fine texture. Their inclusions are mainly fine small-sized quartz grains, small red particles, most likely iron ore, and mica up to 5%. Their feel is smooth and their exterior surfaces are either reddish yellow (5YR 7/6) or light red (2.5YR 6/8). Their fabric matches OXF RS of the National Roman Fabric Reference Collection (Tomber and Dore 1998, 176).

Unknown local fine wares

During the final excavation of the site (AS4143/2002), subsoil layer 3 produced three sherds from a single cup or beaker, 12cm in rim diameter. This vessel is probably local, although its fabric remains unlisted. The fabric is pinkish buff (5YR 8/3) with a distinct grey core. Its fracture is snappy, medium-hard with rough feel. its main inclusions are small fine sub-angular quartz grains, finely distributed in a dense silty matrix, together with few small and rounded black particles, most likely ferrous, and clay pellets. The vessel’s decoration includes two incised rim bands and diagonal combing marks (Figure 2, n.3). Such decoration is likely to imitate early 2nd century AD Colchester colour-coated wares (e.g. Tyers 1996, 167), although such pottery is uncommon in Hampshire.

2.7 Romano-British coarse wares

Flint-tempered wares

Ditch fill 505 in trench 5 (evaluation AS43/2001) produced nine sherds of local flint-tempered wares weighing 234 grams (Plate 3, n.12-20). They come from hand-made vessels, mostly cooking pots and serving bowls, with rough feel and burnished interior surfaces. Their fabric is very coarse, soft and grey (5YR 5/0 or 2.5YR 3/1), containing angular white and grey quartz grains, sparse mica, large angular particles of crushed grey flint up to 25%, occasionally supplemented by large grains of chalk. Such vessels were most likely produced in the New Forest region. The sherds could come from Late Iron Age vessels that were still in use during the Early Roman period; however, it is also likely to represent Late Iron Age potting traditions, which continued after the Roman conquest.

New Forest white wares

The site produced seven sherds of New Forest white wares weighing 52 grams (e.g. Plate 3, n.8-9). Their fabric is the same as the New Forest coarse parchment ware, which was described in section 2.3 for the New Forest mortaria. Such coarse wares are mainly bowls, which have the same pink fracture (2.5YR 7/4) as the New Forest mortaria. Their fabric is medium hard, containing abundant well-sorted quartz grains and sometimes medium-sized angular flint.

Local grey coarse wares (LGCW)

The site produced twenty-nine sherds of local grey coarse wares (LGCW) weighing 506 grams. All sherds come from the New Forest region. Fulford (1975) stresses the similarities between such wares and their equivalents produced by the Alice Holt/Farnham kilns. According to Fulford (1975), it is difficult to distinguish these two fabrics and their identification should be based on a combination of techniques, which include the study of both shape and fabric.

The LGCW fragments recovered from the current site come from hand-made or wheel-made bowls, jars and cooking pots that have been reduced-fired. According to their rim shapes, most come from bowls matching Fulford’s Type 25.2 (1975, 97); however, similar rims are also encountered in Alice Holt/Farnham wares. It is more than likely that such forms were simultaneously produced in more than one fabrics (Lyne and Jefferies 1976). The sherds’ exterior surfaces are rough and most have been wiped hard with a piece of textile, or have been scrapped hard with a rough object. Some sherds are black-burnished, imitating the external appearance of Dorset BB1 wares.

The typical local coarse ware fabric is generally grey (2.5YR 6/0, 7.5YR 6/0) or dark red-brown (7.5YR 4/2, 5YR 3/4), with a wide grey carbon streak covering most of the fracture. It has medium to coarse texture and its hardness varies from medium to soft. The fabric’s main inclusions are medium to large-sized sub-angular to angular quartz grains, small quantities of black and red iron ore particles, mica up to 5% and sometimes grey, white or red clay pellets. Two sherds contain pieces of crushed grey flint. This fabric has no specific reference in the National Roman Fabric Reference Collection; therefore, it is abbreviated as LGCW for the needs of this report. Examples of such fragments are presented in Plate 1 (n.13, 15-16) and Plate 3 (n.21-24). Ditch fill 107 (evaluation AS41/2001) produced a wheel-made rim and neck fragment from a flagon, 9cm in rim diameter (Figure 1, n.7). Ditch fill 505 (evaluation AS41/2001) produced a rim fragment from a late Roman ovoid jar with everted flat rim (Figure 1, n.8).

Alice Holt/Farnham grey wares

The site produced seventy-seven sherds of Alice Holt/Farnham wares weighing 1,221 grams. This material represents 27% of the total ceramic assemblage by count and is the most popular ware recovered from the site. Examples of such sherds are presented in Plate 1 (n.9-10, 14), Plate 2 (n.1-7, 12-13), Plate 3 (n.25-36) and Plate 4 (n.5-6, 10-13, 17-19).

Alice Holt/Farnham wares are reduced-fired and similar to the New Forest grey coarse wares. Their fabric matches ALH RE of the National Roman Fabric Reference Collection (Tomber and Dore 1998, 128). Pottery production in the Alice Holt/Farnham kilns commenced c. AD 50/60 and continued without interruption until the end of 4th century AD (Tyers 1996, 180); therefore, dating such pottery can be challenging. The earlier fabrics tend to be harder and rougher in texture compared to the later ones. Their fabric colour is light grey with darker margins or dark grey throughout. Their exterior surfaces can often be black-burnished, particularly in Dorset ware imitations. After c. AD 200-250, the exterior surface colour of Alice Holt/Farnham wares becomes more ‘metallic’. Fabrics become hard and dense, with smooth texture, decorated in zones with burnished and combed wavy lines (Lyne and Jefferies 1979; Tyers 1996; Tomber and Dore 1998).

Alice Holt/Farnham sherds from the current site vary in relation to their forms and fabrics. The sherds come from both hand-made and wheel-made vessels. Their fabrics are usually grey (2.5YR 6/0, 7.5YR 6/0) or dark brown (7.5YR 3/2) with a distinct dark grey core running along the fracture. The earliest Alice Holt/Farnham fabrics contain medium to large-sized angular quartz grains and various other tempers in sparse quantities, such as clay pellets, red iron ore and very small rounded black particles, possibly ferrous. A single sherd from an early bead-rim form (Plate 3, n.25) comes in a flint-tempered fabric. The latest Alice Holt/Farnham fabrics are tempered with abundant, fine, small to medium-sized sub-angular quartz, in a rather dense and silty matrix. All fabrics contain more than 5% mica, also visible on their exterior surfaces, often producing a sparkling effect. The large quantities of mica found in Alice Holt/Farnham fabrics diversifies them from similar local grey coarse wares (LGCW) from the New Forest region.

The sherds from the current site derive from medium to large-sized bowls, jars and cooking pots. Early vessel forms include globular jars with inverted beaded rims (Figure 2, n.7; Plate 3, n.33), such as the one recovered from ditch fill 505 in trench 5 (evaluations AS43/2001). Similar early Roman forms are noted by Lyne and Jefferies (1979). During the excavation of the site (AS4143/2002) ditch fill 10 produced a single indented beaker sherd and seven flagon sherds. Furthermore, ditch fill 15 produced the rim of a globular jar, 8cm in rim diameter (Figure 2, n.5) and two sherds from an early jar form with beaded rim, 12cm in diameter (Figure 1, n.6).

Dorset black-burnished wares (BB1)

The site produced eighteen sherds of Dorset black-burnished ware imitations which do not match the typical BB1 fabric of the National Roman Fabric Reference Collection (Tomber and Dore 1998, 127). Instead, such imitations are noted in three local fabrics: Hampshire grog-tempered wares (HAM GT), Alice Holt/Farnham wares (ALH RE) and local grey coarse wares (LGCW). In this report, the above eighteen sherds are discussed in relation to their fabric groups. This section is supplementary and is only meant to discuss popular BB1 forms recovered from the site, in relation to their dating.

Original Dorset BB1 wares were hand-made and popular in Southern Britain throughout the Roman period. They were originally produced in the Durotrigian (pre-Roman) tradition of the Middle Iron Age and their production continued throughout the Roman period with various modifications. According to the chronology of some basic forms, flat rimmed black-burnished wares appeared around AD 120, while conical flanged bowls probably developed around AD 250 (Tyers 1996, 182-5). Typical early black-burnished hand-made fabrics are noted as DOR BB1 of the National Roman Fabric Reference Collection (Tomber and Dore 1998, 127), while there is a later BB2 version, associated with the Essex pottery industries.

The two evaluations and one excavation phases at 41-43 Ioannis Anglicus produced body sherds of DOR BB1 imitations, which are difficult to date. However, during the excavation of the site (AS4143/2002), ditch fill 15 produced two LGCW rim sherds, which are distinct. These come from an early Roman ovoid jar form with flat everted rim (Figure 2, n.4; Plate 4, n.15-16), imitating a BB1 ware. This form is closely affiliated with local Late Iron Age - Early Roman potting traditions.

Hampshire grog-tempered wares

The site produced eleven sherds of New Forest grog-tempered wares weighing 177 grams. The sherds come from coarse hand-made vessels, most commonly undecorated jars or bowls, which are occasionally burnished according to the BB1 tradition. Their fabric is dark grey and their main inclusions are grog up to 30%, large particles of angular quartz and sparse mica. The sherds are generally soft and their external feel is rough, with exception of burnished wares, which are smoother. Most of the sherds match fabric HAM GT of the National Roman Fabric Reference Collection (Tomber and Dore 1998, 139). Only few of the New Forest grog-tempered wares are additionally tempered with moderate quantities of local grey flint and chalk. Such fabrics are likely to match fabric LRGR (Late Roman Grog-Tempered Ware), which according to Tyers (1996, 191-2) is likely to associate with Late Roman traditions of the 3rd century AD. Unfortunately, the grog-tempered material from the site preserves no rims or other diagnostic fragments; therefore, any dating based solely on fabric descriptions should be treated with caution.

2.8 Post-medieval and modern pottery

The excavation of the site (AS4143/2002) produced twelve fragments of post-medieval and modern ceramic finds, weighing 644 grams. Eleven sherds were recovered from the topsoil layer and one sherd was recovered together with two Samian fragments from ditch fill 26, which was disturbed. These finds were originally recorded as ‘medieval’; however, their distinct blue or green transfer decoration of flower motifs suggests mass-production techniques of the late 19th and 20th centuries.

Fabric M, imported ink bowls

Topsoil layer 1 (excavation AS4143/2002) produced three intact and identical ink bowls (e.g. Figure 6, n.10). Their base diameter is 5 cm and their rim diameter is 2cm. Their exterior surface is salt-glazed brown (7.5YR 5/5). Their fabric is grey and stone-hard, while its only visible inclusion is sparse mica. All ink bowls carry the same stamped inscription of the word ‘tinteiro’, which is Spanish for ink. Such ink bowls were most likely manufactured in Spain sometime in the 19th or early 20t century.

Porcelain

Topsoil layer 1 and ditch fill 26 (excavation AS4143/2002) produced eight sherds of porcelain with blue/white or green/white transfer decoration. Their fabrics are made of high-fired kaolinitic clay with no visible inclusions. Although the provenance of these sherds could not be identified, they most likely date in the late 19th or early 20th century.

3. Discussion

The discussion of the ceramic assemblage is divided in three sub-sections. Sub-section 3.1 discusses the broader chronology of the site. Sub-section 3.2 discusses stratigraphy and pottery distribution by excavation phase and recovery context. Sub-section 3.2 discusses simplified ware proportions from the site, which are then compared with the results of previous excavations.

3.1 Chronology

Most of ceramic finds recovered during the two 2001 evaluations and the 2002 excavation of the site at 41-43 Ioannis Anglicus Road, date in the Roman period. Modern ceramic finds suggest 19th and/or 20th century activities, which probably disturbed some of the contexts. No ceramic evidence suggests Saxon or medieval occupation of the site.

The earliest Roman evidence from the site is Samian pottery, which dates in the Flavian period (AD 69-96). The same chronology has been suggested by the 1951-4 excavations at Clausentum, where the earliest phases were identified as late Flavian if not Trajanic (Cotton and Gathercole 1958, 12). The presence of flint-tempered coarse wares, Dorset BB1 imitations in local fabrics and transitional forms in grog-tempered fabrics is also likely to suggest a Late Iron Age occupation of the site. Such occupation is represented by earlier ceramic traditions, which most likely continued during the Roman period. In the early excavations at Clausentum, Cotton and Gathercole (1958, 12) recorded similar evidence, which confirmed some Iron Age occupation on site.

According to the most recent Roman ceramic evidence, fabrics and forms suggest a continuous occupation between the Flavian period and the late 4th century AD. This chronological span is confirmed by the presence of New Forest mortaria dating between AD 260-370, late Roman grey coarse ware forms (LGCW), late Roman Alice Holt/Farnham wares with distinct ‘metallic’ surfaces, and perhaps, 3rd to 4th century AD grog-tempered fabrics from the New Forest region, which are additionally tempered with coarse flint.

3.2 Stratigraphy

Based on the pottery from each context and the Harris Matrix, which is submitted with the original site report, it is difficult to discuss chronological phases and context associations. The trenches that were dug during the two evaluations in 2001 suggest that most of the fills have been re-deposited, as they contain pottery of mixed early and late Roman dates. This re-deposition of the Roman finds is likely to had followed the construction of the Saxon cemetery in 75 Ioannis Anglicus Road, although other evidence would have suggested direct associations. So far, no traces of Saxon burials or other Saxon activities have been noted within the plot at 41-43 Ioannis Anglicus Road. The analysis below discusses the chronological distribution of the ceramic material by trenches and contexts, explaining the degree its re-deposition:

Evaluation AS41/2001

Trench 1: Dressel 20 and Gauloise 4 amphorae were present in the four upper fills of the same ditch (113, 119, 103, 116), while Gauloise 4 fragments continued further down in ditch fills 125, 126 and 114. Dressel 20 amphorae were mainly produced between c. AD 1-250 while Gauloise 4 between c. AD 30-250 (Tyers 1996, 86). Such evidence suggests that the upper and middle fills of the ditch date between the 1st and 3rd century AD. This is further confirmed by the presence of early Alice Holt/Farnham wares, found in the same contexts. By contrast, ditch fill 107, which is the lowest ditch fill, produced pottery of diverse chronology: Flavian Samian (c. AD 70-100) was found together with New Forest brown-slipped wares (c. AD 260-370) (Tyers 1996, 166) and late Alice Holt/Farnham ‘metallic’ wares of the late 3rd-4th century AD (Tyers, 1996, 180). These observations suggest that the content of the ditch evaluated in trench 1, was probably re-deposited sometime in the late Roman period.

Trench 2: single fill 209 of pit 210 was chronologically homogeneous, containing late Roman fabrics. However, fill 212 of pit 213 produced Oxfordshire red-slipped wares dating between c. AD 240-420 (Tyers 1996, 166), mixed with Dressel 20 amphorae fragments (c. AD 1-250 AD). This suggests that pit 213 was filled sometime in the late Roman period with mixed debris, which included earlier fragments. By contrast, single fill 216 of pit 217, which was dug under pit 213, produced a combination of Flavian Samian sherds, local New Forest wares produced after c. AD 250 and Gauloise 4 amphorae (c. AD 30-250). This observation is likely to suggest that pits 213 and 217 represent complicated re-redeposition events, taking place during or right after the late Roman period.

Trench 3: ditch terminal fills 303 and 305 produced limited ceramic evidence, which are not informative of date and provenance; therefore, this material can be characterised as broadly Roman.

Trench 4: Ditch fills 405 and 409 produced sherds of mixed chronology, such as Gauloise 4 amphorae (c. AD 30-250) and possibly late Hampshire grog-tempered wares (c. 360-420 AD) (Tyers 1996, 180). The lowest ditch fill 409 contained local grey coarse wares with no distinct chronological characteristics. It seems likely that the ditch was filled up during the late Roman period with Roman debris dating across four centuries.

Evaluation AS43/2001

During this evaluation phase at 43 Ioannis Anglicus Road, five trenches were dug on site, of which one produced pottery. Trench 5 revealed a single ditch consisting of eleven fills; pottery was only found in two of its middle layers. Ditch fill 506 contained late Roman New Forest Mortaria dating c. AD 260-370 (Tyers 1996, 117) and a single sherd of Samian that is highly deteriorated after long deposition and possible re-deposition. Ditch fill 505 directly below ditch fill 506, contained early and late Alice Holt/Farnham wares, local grey coarse wares, Dorset BB1 imitations, a late Roman New Forest mortarium, Flavian Samian wares and flint-tempered wares associated with Late Iron Age traditions. Based on the pottery, the entire ditch was probably re-deposited towards the end of the Roman period.

Excavation AS4143/2002

During the excavation of the site, topsoil layer 1 produced post-medieval/ modern ceramic finds, which included Spanish ink bowls and various porcelain fragments. Subsoil layer 3 produced Gauloise 4 amphorae fragments (c. AD 30-250) and a possible local imitation of an early 2nd century AD Colchester colour-coated ware. Further excavation revealed a single drainage ditch, which was dug in two segments, divided in a total of four fills. In the first segment, ditch fill 10, which represented the upper layer of the ditch, produced possible late Roman Hampshire grog-tempered wares and late Alice Holt/Farnham wares, mixed together with Dressel 20 amphorae fragments (c. AD 1-250). Ditch fill 15, directly underneath ditch fill 10, produced Dressel 20 amphorae fragments (c. AD 1-250), early and late Alice Holt/Farnham wares and late local grey coarse wares. In the second segment, ditch fill 26, which represented an upper fill, produced Samian sherds of Flavian date, mixed with modern porcelain sherds. Ditch fill 23, directly beneath ditch fill 26, produced Samian wares of Flavian date, Gauloise 4 amphorae fragments (c. AD 30-250), late Roman Alice Holt/Farnham wares (late 3rd-4th century AD), New Forest brown-slipped wares (c. AD 260-370) and Hampshire grog-tempered coarse wares of possibly early and late Roman dates. The analysis of the material from the four ditch fills suggests that the ditch was re-deposited sometime in the later Roman period.

3.3 Ware distribution

Chart 1 presents the distribution of the ceramic assemblage by count and weight, divided in six simplified ware categories. According to the chart, almost half of the total ceramic assemblage by count consists of Romano-British coarse wares, followed by almost a quarter of imported coarse wares. This means that 74,4% of the total ceramic assemblage consists of coarse utilitarian pottery and only 21.4% associates with local and imported fine ware fabrics. Even though the majority of coarse wares are locally made, the majority of fine wares (14.4% by count) are imported from the European mainland.

If the same assemblage is examined by weight, fine wares appear in small percentages due to their low weight. Furthermore, imported coarse wares appear in high weight percentage due to some heavy amphora sherds, which bias the statistics. Again, the analysis of the assemblage by weight suggests that 84.8% of the pottery comes from utilitarian coarse wares and only 7.5% associates with Roman fine wares. In general, the analysis of the material by both count and weight suggests similar percentages.

Chart 1: The total assemblage divided in simplified ware categories

By comparison with the pottery published by Cotton and Gathercole (1958) from the 1951-4 excavations at Clausentum, the current assemblage is characterised by the same fabric divisions. The main difference between previous and current excavations is that the latter did not produce any Gallo-Belgic pottery (Terra Rubra and Terra Nigra), while Late Iron Age ‘Romanised’ wares were limited. Of course, the area of the recently excavated site is significantly smaller compared to the area excavated in 1951-4; therefore, the current material does not represent the total of the Roman activity on site. Furthermore, most of the sherds included in this report are non-diagnostic body fragments and cannot assist in a full discussion of vessel forms and relevant dating.

In the absence of such information, the discussion of pottery consumption patterns from the site does not add much to the older analysis by Cotton and Gathercole (1958). The quantification methods in this report suggest the prevalence of local coarse wares, while fine wares appear to be mainly elite imports. Such conclusions match Tyres (1996) and Fulford (1975), who suggest that local fabrics are the most popular in Roman sites across Southern Britain, particularly during the late Roman period.

Bibliography

Cotton, M.A. and Gathercole, P.W., 1958, Excavations at Clausentum, Southampton 1951-1954, London: Her Majesty’s Stationery Office.
Fulford, M.G., 1975, New Forest Roman Pottery, Oxford: British Archaeological Reports Series 17.
Lyne, M. and Jefferies, R.S., 1979, The Alice Holt/Farnham Roman Pottery Industry, London: Council for British Archaeology Series 30.
Tomber, R. and Dore, J., 1998, The National Roman Fabric Reference Collection: A Handbook, London: Museum of London Archaeology
Tyers, P.A., 1996, Roman Pottery in Britain, London and New York: Routledge.

Figures

Figure 1: Roman mortaria, amphorae and bowls, including modern finds

Figure 2: Late Iron Age to Roman coarse wares and Samian imports

Plates

Plate 1: Evaluation AS41/2001, trench 1, ditch fill 116

Plate 2: Evaluation AS41/2001, trench 1, ditch fill 118

Plate 3: Evaluation AS43/2001, trench 5, ditch fill 505

Plate 4: Excavation AS4143/2002, ditch fill 15