Notes on the Dating of Late Shell-Tempered Ware in Essex

Colin Wallace

Introduction: The evidence from Chignall St lames

In the late 1970s and early 1980s, several seasons of excavation took place in advance of quarrying to the south of the known Roman villa at Chignall St James in Central Essex (TL662108: some five km. north­west of Chelmsford). In the forthcoming report, several later Roman wares are relied on quite heavily for the purposes of relative dating. The excavator, when discussing the dating evidence from components of site Periods Ill. 3 (c AD245-285) and Ill. 4 (c AD 285-370), had suggested inception-dates for four late Roman wares as part of this scheme, based on the site-sequence at Chignall (Clarke forthcoming) .

For Hadham oxidised ware, Rettendon wares and Oxfordshire red colour-coat, the dates derived from the stratigraphic sequences accord with those widely­accepted and are not further considered here. However, the contention that "late Roman shell­tempered ware first appears on the site in the earlier 4th century" (my italics) may seem somewhat surprising (compare with Going 1987,118).

Put simply, the observed evidence from Chignall is that, in Period III 4, late shell-tempered ware occurs in the primary silt of the system of interconnected enclosures or paddocks (Enclosures I-L) , as well as in fills belonging to the later backfilling episodes of Stratigraphical Zones Band A, while Oxfordshire red colour-coat occurs later in the sequence. Again, with the complex of folds within these paddocks (the subrectangular enclosure complex), late shell-tempered ware (henceforth referred to as LSH) occurs in fills earlier than Zone A (c AD370, when Oxfordshire ware appears). The obvious interpretation to put on this is that LSH is present rather earlier than Oxfordshire red colour­coat, breaking with the later date and loose

123

association between the two wares established at, for example, nearby Caesaromagus (Going 1987, 115 and 118). The coin (Gloria Exercitus, one standard) in Zone B helped to suggest the inception-date for LSH quoted above.

The accepted dating

In Essex, late shell-tempered ware has been seen as appearing c AD360/370 (eg Drury 1977, 45). Higgins' Caistor on Sea Group 3 (with coins of the House of Valentinian, AD367-375) may have been a major influence - although earlier 4th century stratigraphy was absent on this site (Higgins 1973, 286-299). The last general survey of LSH (unpublished BA dissertation, Institute of Archaeology 1973, by Janet Sanders: consulted by kind permission of Professor J. J. Wilkes) could find no published examples of LSH jars (her form 3 =

Going G27) earlier than c AD350. A recent study of later Roman pottery groups in southern Britain could find no evidence to controvert the later 4th century date for the widespread distribution of LSH in quantity (Pomel 1984, 107-113). Recent work by Going on cycles of pottery production has suggested, also, that the period c AD300-350 has yielded few good groups and can only be characterised as "one where certain forms and products are not yet evident" rather than one where new wares were being developed and industries were flourishing (1992,101).

A survey of the literature

In this light, the suggestion made by Clarke that LSH appears on this site in the period c AD300-350 would seem to be at variance with the generally­accepted dating. A brief survey of the recent literature, below, does provide some support,

124 Colin Wallace

although it is largely confined to sites some way to the west of central Essex. Regrettably, when studying the pottery from the crucially-positioned site at Ivy Chimneys, Witham (Turner-Walker and Wall ace forthcoming), the suggested redating was taken at face-value rather than being tested against the site-sequence.

In the absence of a decent corpus of the vessels associated with Late Roman coin-hoards, it is difficult to take advantage of the useful fixed points that these pots would provide: I note however a Going G27 jar from Caister-on-Sea (Darling 1993, fig. 150. 462) which held a hoard with latest coin of AD340 (Reece, in Darling 1993, 62) .

An important point to note is that what we in Essex call 'Late shell-tempered ware' is merely a facet of a long-lived East Midlands tradition (eg Perrin and Webster 1990, 40ft). Some southern sites are receiving these wares from the mid-1 st century AD onwards (eg Verulamium: Wilson 1984a, fig. 95. 2283) and while in North Wales the ware is chiefly a late 4th-century+ one (Webster 1993, 254), at CaernarfonlSegontium there was also some 2nd­century material, VER 666-type jars (Webster 1993, nos. 172,332 and 381) .

Sites with LSH present c AD360+ are as follows:

London - Angel Court. From the site-sequence of deposits on the bank of the Walbrook (Orton, in Blurton and Rhodes 1977, 43), LSH (and Oxfordshire) first occurred in a layer with latest coins AD 364-378. - Tower of London, Inmost Ward . The group dumped against the late 4th-century second riverside wall had LSH, Oxfordshire red colour­coat and latest coin AD 388-392 (plus many of 364-378) (Cameron, in Parnell 1988, 57) .

Caister on Sea (above) (The report on the 1950s excavations (Darling 1993) is less than keen on applying absolute, as opposed to relative, dates to the site-sequence, for good reasons) .

Enfield - Lincoln Road. The Phase 4 ditch-fill group (Tyers, in Gentry et al 1977, nos. 21. 1-16) had LSH, Oxfordshire, Hadham ware with 'Romano­Saxon' decoration and coins of AD 364-378.

Chelmsford - Going (1987, 51-58) Groups 18 and 21.

Kelvedon - Area J cemetery. The base of an LSH jar was used to cover a Hadham face-pot, a type dated c AD 340-400 + (pers in! Chris Going), in grave 67 (Rodwell 1987, 119) .

Thetford - Fison Way. The Late Roman group from the uppermost fill of the Late Iron Age/early Roman enclosure (Gregory 1992, 172) had an

LSH jar, Oxfordshire, Nene Valley, greywares with 'Romano-Saxon' decoration and a coin of AD 364-378 (probably).

LSH appears in contexts of AD 300-350, largely on sites to the west of the modern county, at:

Brentford (site 9) - Jars in groups Z2 and Zl, along with Oxfordshire red colour-coat, Hadham oxidised wares and Nene Valley colour-coat. Latest coins AD 324-326 and 337-341 respectively (Laws 1978, fig. 50. 65 and 49. 27/28).

Old Ford - Jar in a first half 4th century pit (McIsaac, in McIsaac et al 1980, fig. 18. 112) as well as in a late 4th century ditch.

Sewardstone - LSH jars present in contexts with Oxfordshire red colour-coat and a coin of AD 330-335 (Huggins 1979, 185).

Braughing - 1971-72 excavations. Key Group D22, latest coin AD 330-335, contained an LSH jar (Potter and Trow 1988, fig . 61 . 229) .

Baldock - From the series of stratified groups (Stead and Rigby 1986, 273-379), the distinctive rilled LSH jars first occur in the late 3rd/early 4th century well A483 (fig. 154.726-728) and are first associated with Oxfordshire ware in a mid-4th century group.

Verulamium - 1955-61 excavations. LSH jars have date-ranges beginning in this period (Types 1283 (Wilson 1984b, 276: AD 310-410); 1871 (Wilson 1983,339: AD 340-380) and 2194 (Wilson 1984a, 223: AD 350-400» while the bowls are seen to be late (Types 1258 (Wilson 1984b, 276: AD 360-410+) and 2425 (Wilson 1984a, 243: AD 360-400» . - King Harry Lane. On this site in the south­western extramural area,a group dated to the early 4th century (from a pit cutting the surface of the Silchester road) contained an LSH storage-jar, along with some possible Oxfordshire red colour­coat, Nene Valley colour-coat and BB1 (Stead and Rigby 1989, 77). There were no coins, and the date was not argued-for: perhaps it was influenced by the proposed later 3rd-century desertion of the extramural area (Stead and Rigby 1989, 11): there were no coins later than three copies of the 350s/360s from the whole site (Reece, in Stead and Rigby 1989, 13).

At the Saxon Shore fort of Burgh Castle, Norfolk, Johnson (1983, 27) was able to isolate a possible destruction-deposit belonging to the main occupation coin-dated to the 340s (Hammerson, in Johnson 1983, 69). LSH was present, along with Oxfordshire

Colin Wallace 125

colour-coat. However, the coin-dating may only be providing a terminus post quem.

Coin-dating of 4th-century deposits

There were gaps in the supply of regular 4th century coinage (in the 340s and 350s) but these were more than made up for by widespread copying. However, coin-dating of deposits in eastern Britain between c AD330 and 360/370 may be a problem. For this critical period, decent sequences of coinage are lacking. This shortage was first noticed by Reece when commenting on the coins from the 1968-73 excavations at Kelvedon: it and other small settlements in the east of Britain showed themselves low in late coin loss (coins struck between AD 330 and 402) by comparison with small settlements elsewhere in the province (Reece, in Rodwell 1987, 80-81 and table 6; Reece 1991, 102-103) .

The problems of coin-dating late Roman deposits were reflected on by Graham Webster (1977 , 321-322), who pointed-out that only 1.4% (thirty­one coins out of 2192) of the coinage in the late fourth-century Verulamium Theatre deposit established its date. More than half of the coins were issues of the House of Constantine, perhaps still in circulation. The coins from Period 5 at Orton Hall Farm in the lower Nene Valley (dated c AD 350/375 +) were of the later 3rd century and the AD330s, as well as the 360s-370s and the 390s (pers comm 1 R Perrin).

Assemblages of actual later date may therefore only have Constantinian t.p.q.'s - a case in point is period V. 2 at Heybridge, where LSH and Oxfordshire colour-coat occur with coins of the 320s and 340s (Wickenden 1987, 17 and 22) . Preliminary study of the coins accompanying three deposits with LSH on a site in North-West Essex has shown them to be of Constantine I (AD 318-324) in one group and of the House of Valentinian (AD 360s/370s), along with a copy of the 350s/360s, in another. From the pottery evidence, the groups all appear contemporary. Another case may be the material from Burgh Castle (above) . This site too has a shortage of coinage over the critical period and the tentative coin-dating is at variance with some of the pottery (eg the Oxfordshire colour-coat and the Hadham sherds with 'Romano-Saxon' decoration). The early 5th-century glass hoard (Harden, in 10hnson 1983, 81-88) was deposited slightly later in the site-sequence.

Endword In summary, with regard to the occurrence of late shell-tempered ware on Essex sites, what is certain is that the ware is present in quantity in the 360s and later, what is probable is that the inception-date is earlier than this and what is possible is that this inception-date may be some thirty years earlier than previously suggested. In view of the shortage of coinage mentioned above, the present writer is less sure about this last suggestion than the excavator of Chignall and his original pottery reporter - and so I offer these notes for discussion, rather than as a definitive statement.

Colin R. Wallace 1991, revised 1992 and 1993.

Bibliography Blurton, T. R. and Rhodes, M. 1977: 'Excavations at Angel

Court, Walbrook, 1974', Trans London Middlesex Archaeol Soc 28, 1977, 14-100.

Canham, R. (ed) 1978:, 2000 years of Brentford, London. Casey, P. J. , Davies, J. L. with Evans, J. 1993: Excavations

at Segontium (Caernarfon) Roman Fort, 1975-1979, Counc Brit Archaeol Res Rep 90.

Clarke, C. P. forthcoming: 'Excavations to the south of Chignall Roman villa, Essex 1977-81', East Anglian Archaeol.

Darling, M. J. with Gurney , D. 1993: Caister-on-Sea. Excavations by Charles Green, 1951-55, East Anglian Archaeol60.

Dore, J. and Greene, K. (eds) 1977: Roman Pottery Studies in Britain and Beyond, Brit Archaeol Rep S30.

Drury, P. J 1977 : 'Braintree: Excavations and Research, 1971-76', Essex Archaeol Hist 8, 1976, 1-143.

Frere, S. S. 1983: Verulamium Excavations Il, Rep Res Comm Soc Antiq London 41 (London).

Frere, S. S. 1984: Verulamium Excavations Ill, Oxford Univ Comm Archaeol Mon 1 (Oxford) .

Gentry, A. , Ivens, J. and McClean, H. 1977: 'Excavations at Lincoln Road, London Borough of Enfield, November 1974-March 1976', Trans London Middlesex Archaeol Soc 28, 1977, 101-189.

Going, C. J. 1987: The Mansio and other sites in the south­eastern sector of Caesaromagus: the Roman pottery, Counc Brit Archaeol Res Rep 62 (London).

Going, C. J. 1992: 'Economic "long waves" in the Roman period? A reconnaissance of the Romano-British ceramic evidence' , Oxford J Archaeol 11, 1992, 93-117 .

Gregory, T. 1992: Excavations in Thetford, 1980-1982, Fison Way , East Anglian Archaeol53 .

Higgins, D. C. 1972: 'Three Groups of Romano-British Coarse Pottery from Caister-on-Sea, Norfolk', Norfolk Archaeol35 , 1970-73,279-301.

Huggins, R. M. 1979: 'Excavation of a Late Roman Site at Sewardstone Hamlet, Waltham Holy Cross, Essex, 1968-75', Essex Archaeol Hist 10, 1978, 174-188.

126 Colin Wallace

Johnson, S. 1983: Burgh Castle, Excavations by Charles Green 1958- 61, East Anglian Archaeol20.

Laws, A. J. 1978: 'Roman coarse pottery', in Canham 1978, 42-75.

Mclsaac, W., Schwab, I. and Sheldon, H. 1980: 'Excavations at Old Ford, 1972-1975', Trans London Middlesex Archaeol Soc 30, 1979, 39-96.

Parnell, G. 1988: 'The Roman and Medieval Defences and the Later Development of the Inmost Ward, Tower of London: Excavations 1955-77', Trans London Middlesex Archaeol Soc 36, 1985, 1-79.

Perrin, J. R. and Webster, G. 1990: 'Roman Pottery from Excavations in Normangate Field, Castor, Peterborough, 1962-3', J ROrn£1n Pottery StuLl3, 1990, 35-62.

Pomel, M. G. 1984: 'A Study of Later Roman Pottery Groups in Southern Britain: fabrics, forms and chronology', unpublished MPhil. thesis, Univ Coli London Inst Archaeol.

Potter, T. Wand Trow, S. D. 1988: Puckeridge-Braughing, Hertfordshire: Ermine Street Excavations 1971-1972. The Iron Age and ROrn£1n Settlement, Hertfordshire ArchaeollO, 1988.

Reece, R. 1991: ROrn£1n Coins from 140 Sites in Britain, Cotswold Studies IV (Cirencester).

Rodwell, K. A. 1987: The prehistoric and ROrn£1n settlement at Kelvedon, Essex, Counc Brit Archaeol Res Rep 63 (London).

Stead,1. M. and Rigby, V. 1986: Baldock. The Excavation of a Roman and Pre-Roman Settlement, 1968-72, Britannia Mon 7 (London).

Stead, I. M. and Rigby, V. 1989: Verulamium: the King Harry Lane site, English Heritage Archaeol Rep 12.

Turner-Walker, C. and Wallace, C. forthcoming: 'The Iron Age and Roman Pottery', in Turner, R, Excavations of an Iron Age settlement and Roman religious complex at Ivy Chimneys, Witham, Essex 1978-83', East Anglian Archaeol.

Webster, G. 1977: 'Reflections on Romano-British pottery studies, past, present and future', in Dore and Greene 1977, 317-333.

Webster, P. V. 1993: 'Coarse pottery', in Casey et al1993, 250-309.

Wickenden, N. P. 1987: 'Prehistoric Settlement and the Romano-British 'Small-town' at Heybridge, Essex', Essex Archaeol Hist 17, 1986,7-68.

Wilson, M. G. 1983: 'The Pottery Used for Dating', in Frere 1983,294-34l.

Wilson, M. G. 1984a: 'The Other Pottery', in Frere 1984, 200-263.

Wilson, M. G. 1984b: 'Revised Dates of Pottery Published in Volume 1' , in Frere 1984, 267-276.