15

The region and its resources

The estuary

London made significant demands on regional resources: both for its own needs and for onward supply. The establishment of the city spurred an expansion of extraction industries (Fig. 15.1). The most archaeologically visible include the stone quarries of northern Kent, salterns along the Thames estuary, iron working in the Weald, and woodlands that supplied London’s fuel and building timber. These existed alongside the agricultural exploitation of arable lands, principally for grain, and pastures for livestock providing traction, meat and secondary products such as leather. Cattle bones form a larger part of assemblages from the London basin compared to the rest of south-east Britain where sheep/goat are more common. ¹ This may be because of the importance of plough animals in an area where grain was an important crop, and the use of oxen for haulage in and around London. Horse bones are also found in greater abundance on rural sites in the Thames estuary and London basin, reflecting on the development of cattle-droving and horse-breeding for the Roman cavalry.

Fig. 15.1 Some of the key resources exploited by London and the Roman administration in south-east Britain, chiefly in the late first and early second centuries. Drawn by Fiona Griffin.

London has no building stone and the Roman city imported Kentish ragstone for most of its concrete constructions. This sandy grey limestone was quarried from the banks of the river Medway around Maidstone and came to be used in Roman constructions within a decade of the conquest. ² Its export may have contributed to the precocious development of a villa economy in this part of Kent. ³ This in turn suggests that local land-owning aristocrats profited from their ownership of these mineral resources. ⁴ Whilst the stone may have been acquired from private estates, quarrying and transport are more likely to have been undertaken by licensees managing the exceptional demands of large public building projects. Ragstone only came into widespread use as a building material in London from c. ad 80, involving regular shipments into the Thames and up-river to London’s quays (below p. 308).

The estuary wetlands provided a wealth of resources. ⁵ Watermeadows in the Thames and Lea valleys were used for pasture: supporting oxen needed for traction, animals destined to be slaughtered for their meat and hides, equids for official use, and haymaking for winter fodder (below p. 193). Cattle were herded to town for slaughter, passing through roadside sites such as Old Ford, Staines and Enfield. Some livestock was butchered at these places, but most reached London on the hoof, illustrated by carcass processing waste from the city. ⁶ Roman intervention in cattle management is suggested by stock improvements evident in animal bone assemblages, as larger breeds of cattle were introduced to the south-east. ⁷ Investment on this scale might have been easier to achieve on larger managed estates, and allowed for the breeding of the stockier and more muscular cattle useful for traction. Sheep would also have been pastured on down heathlands and estuarine saltmarshes.

The river offered opportunities for fishing and fouling, and oyster beds were exploited on the Essex and Kent coasts. ⁸ Products reached London by river, and dumps of oyster shells around the late first-century landing stage at Peninsular House suggest that this was where they arrived. ⁹ The value of many wetland resources was realized through the use of salt to transform perishable seasonal produce into commodities transported to towns and forts for over-winter consumption. ¹⁰ Salt enabled taxation. James Gerrard has estimated that if the entire meat ration of the Roman army needed curing, this alone would have required some 113 tons of salt annually. ¹¹ The potential to extract salt from evaporating tidal pools was enhanced by digging brining channels, followed by boiling the product over hearths in briquetage vessels. This was a seasonal activity, best undertaken in the early autumn or later spring. Red hills, formed of waste from burning marsh plants to produce salt ash to strengthen the saline content of the brine, are found on both sides of the Thames estuary, as at Canvey Island, on the Upchurch marshes and the Hoo Peninsula. Following rapid early Roman expansion this production peaked in the first and second centuries AD. ¹²

Salt production was accompanied by pottery production. Containers were needed to store salt and salted goods, and a shared dependency on fuel and river transport added to the utility of establishing winter potteries at the sites of salterns. James Gerrard has suggested that the manufacture of Black Burnished ware (BB1) in Poole Harbour in the first century AD was associated with local salt production for military supply. ¹³ This seems likely to have involved the military control of production, perhaps involving requisitioning of salt and salted goods as a form of taxation. Similar arguments can be advanced for the pottery industries established within the Thames estuary, including early Roman lid-seated shell-tempered jars and later production of black burnished pottery. ¹⁴ The army was probably the most important market for these goods, but London is likely to have managed their procurement. The estuary wetlands attracted relatively few villas and high-status settlements, with a material culture that was poorly integrated with areas to north and south. ¹⁵ Some of these territories may have comprised estates administered from London, possibly including imperially owned land.

Woodlands

London relied on woodlands for its fuel and building materials. We have already described the importance of timber in developing the urban infrastructure, especially along the waterfront. A further supply of structural timber went into the walls, roofs and floors of London’s houses, with more destined for shipyards. The heated public baths had massive fuel requirements, alongside industrial and domestic use. It was prohibitively expensive to transport wood by road, and most would have reached London by water. Oak was favoured for most uses. The timbers used in London present similar tree-ring profiles, suggesting that they were sourced sufficiently close to the city to have grown in like conditions. Extensive areas of coppiced woodland were found on the London claylands where soils are ill-suited for agriculture. The relative scarcity of settlement sites and field ditches suggests that such areas retained a significant degree of tree cover. Woodlands upriver of London would always have been preferred for ease of transport. As in the medieval period, London probably drew on woods in Surrey, Hampshire, Berkshire and Hertfordshire. ¹⁶ We do not know on whose land the trees were grown, although woods could be privately, publicly and imperially owned. ¹⁷ It is probable, however, that most procurement was undertaken against specific project needs. Timber was generally only felled when required, and supervising carpenters probably visited the woods to select the trees. ¹⁸

Damian Goodburn’s studies of structural timbers give detailed insight into forestry management around London. ¹⁹ Coppiced timber and underwood was harvested from managed woodland, on anything from 3- to 40-year rotations, with an emphasis on producing small, fast-grown timbers for house-building. For example, the timbers used in a warehouse dated ad 152–3 found at the Courage Brewery site in Southwark were obtained from oak coppice cut on a long rotation, witnessing extensive woodland management that started no later than the late first century. ²⁰ These managed woods existed alongside areas of wildwood, where larger slow-grown and straight-grained timbers were found, pockets of which were still available for exploitation into the third century. Additional supplies were obtained from field and hedgerow oak, and alder from the riverbanks.

The felling and transport of London’s timber was a labour intensive process, best undertaken in the autumn and winter when lower moisture content reduced the weight of timber needing transportation. ²¹ It was also easier for carpenters to select the trees suited for their needs when leaves had fallen. After felling, unwanted side branches were removed and the main stems cross-cut or ‘bucked’ as necessary. This was accomplished using thin-bladed axes similar to the Roman army dolabra. The large oak baulks used in the waterfront quays were probably dragged by oxen to collection points on the banks of the Thames and its tributaries to be loaded into barges. These timbers could be over 6 metres long and weigh up to two tonnes, making them too heavy to float unaided. Conversion into beams, planks, and boards is likely to have taken place in town, but trimmed branches and other logging debris would have been converted into charcoal.

The ready availability of such fuel supported industrial uses bordering forested areas. This, combined with the availability of clay and excellent transportation contributed to the development of pottery kilns along Watling Street, between Brockley Hill and Verulamium. ²² These made a standardized series of specialist forms not found in the repertoire of native potters, including flagons and mortaria in oxidized fabrics. Verulamium region wares are found in archaeological contexts at One Poultry predating ad 53–5 and production probably commenced before ad 50. ²³ By c. ad 60 this industry provided nearly a quarter of all oxidized wares used in London, and more than three quarters of such wares by the end of the century. The rapid emergence of this industry relied on immigrant potters who set up in business to supply urban and military consumers. The kilns located close to the posting station at Sulloniacis might have suited the interests of a lessee supplying public and military contracts before urban markets were established. An opportunity of this sort might account for the early date of settlement at Brockley Hill, and explain the distribution of these products along army supply routes. ²⁴ Mortaria and flagons in Verulamium region white-ware were widely used along the northern frontier in the Flavian-Trajanic period. ²⁵ Intriguingly the early products of these kilns included amphorae of a type (Dressel 2–4) that usually contained wine, suggesting that vineyards were established near London by c. ad 60. ²⁶

Most of the utilitarian coarse wares used in early Roman London were made in potteries at Highgate Wood, between Watling Street and Ermine Street, some 10 kilometres north-west of town. Ten kilns and associated features were excavated here in the late 1960s and early 1970s. ²⁷ From c. ad 50/55 onward jars and bowls were made in the pre-Roman ‘Belgic’ style using grog-tempered fabrics. Early in the Flavian period, however, new and more efficient kilns started producing wheel made Romanized sand-tempered wares. By the end of the century, native pottery traditions had all but disappeared. ²⁸ The kilns were located near to the woods that fuelled them, and it is likely that they took advantage of the transport links and seasonal labour brought here by woodsmen and charcoal burners. This relationship might help explain the success not only of the Highgate Wood pottery industry, but also of London’s import of Alice Holt pottery from near Farnham in Surrey. ²⁹ These potteries on the borders of woodlands produced grey-ware jars, dishes and bowls soon after the Roman conquest, that were abundant in London down to c. ad 90 (and again after c. ad 270). They probably reached London along supply routes that followed the navigable rivers Slea and Wey down to the Thames. The Farnham area may have been an important source of London’s building timber, as it was in 1395 when it supplied timber for the roof of Westminster Hall. ³⁰

The forested claylands around London also supported brickworks that supplied London with tiles for hypocaust systems, roofs, and floors. Around 90 per cent of the tile used in London in the late first and early second centuries came from kilns within 30 kilometres of the city. ³¹ These were usually near sites involved in supplying London with other bulk materials, exploiting the same transportation infrastructure. Some tile was imported from parts of Kent involved in the supply of Kentish ragstone along the Medway and Thames. Another tilery supplying London was located on the south coast of Sussex: products are likely to have reached London overland following routes established to transport Wealden iron, and from the area of potteries at Ashtead (Surrey) and Brockley Hill. ³² London also imported pottery from North Kent, joining the carriage of grain and ragstone down the Medway and up the Thames.

Wealden iron

The metalled roads that penetrated the Sussex Weald were the likely product of military engineering during the early Flavian reorganization of the transport infrastructure around London. ³³ We have already noted that Stane Street’s entrance to London was laid out in the early Flavian period. A similar date obtains from roadside ditches at the southern end of the London-Lewes road, at Bridge Farm. ³⁴ Two of these roads traversed the High Weald, a notoriously intractable and sparsely occupied wooded terrain, opening up the area for the export of iron. Over 100 Roman iron-smelting sites are known from the Weald, with many clustered against the roads to London. ³⁵ Although the earliest iron working pre-dated the conquest, the industry expanded rapidly after the invasion, contributing to a fourfold increase in the number of sites and even larger increases in volumes of production. Some witnessed exceptionally high levels of industrial activity, although much production remained small-scale, probably integrated into the annual farming cycle and drawing on a skilled local workforce. In addition to producing the iron needed for nails, chains, and fittings, the Weald may have provided ship-building timber. ³⁶

Iron blooms would have been brought into London by ox-drawn cart, supplying smiths working in Southwark. ³⁷ The growth of this industry was stimulated by the value placed on iron. Texts found at Vindolanda on Hadrian’s Wall show that refined iron commanded a high price, perhaps about 1.1 denarii per kilogram in the early second century. ³⁸ A massive slagheap associated with a major production site at Beauport Park, just north of Hastings, suggests that production here averaged 50–65 tons of iron a year, with an estimated value of up to 71,500 denarii. At the close of the first century, this was equivalent to the annual salary of 317 legionary soldiers, fully covering the costs of extraction and transportation. The iron produced at Beauport Park is unlikely, however, to have supplied London. Output from the south-eastern Weald was more sensibly taken to the coast for the Roman fleet, Classis Britannica, through ports on the English Channel. Tile kilns operating alongside the iron smelting manufactured tegulae carrying the CLBR stamp of the Classis. Massive numbers were used in the construction of the bathhouse at Beauport Park, where a fragmentary inscription recorded repairs possibly undertaken by a vilicus named Bassus or Bassianus. ³⁹ Henry Cleere has argued that the distribution of CLBR stamped tiles defines a maritime area controlled by the fleet deep into High Weald. ⁴⁰ He proposed that iron-production was under direct procuratorial control, with operations handled by civilian workers supervised by the Imperial vilicus and integrated into the supply-chain controlled by the fleet. ⁴¹

During the reign of Antoninus Pius, the post of procurator was, on at least one occasion, combined with command of the British fleet. ⁴² This united control of key raw materials, including the iron and timber extracted from the eastern Weald, with the enhancement of the naval capacity that relied on these materials. In the absence of evidence for a significant distribution of villas or other elite settlements, Cleere proposed that the Weald remained part an Imperial estate managed by the procurator. A medallion dating to the reign of the emperor Antoninus Pius found at Bardown had probably been given to a very high-ranking official, reflecting on the importance of this command. ⁴³

Production in the western high Weald may have been differently organized. ⁴⁴ Sites in this area were not supplied with tiles produced by the Classis, and relied on the road links to London for their export. One possibility is that the iron smelting operated through concessions leased to middlemen (coloni), who paid a share of production to the Imperial fiscus. ⁴⁵ It is important to note that these differences in approach to managing surplus exaction don’t imply differences of tenurial arrangements or in the state’s interest over the iron produced. ⁴⁶ It is perfectly possible that some parcels of land, imperially owned or not, were managed by leaseholders supplying London, whilst others were farmed by public officials to secure supplies for Channel ports operated by the fleet. In both arrangements the ironworkers might have been found locally, whether paid for their services, organized as share-croppers, or as forced labour. It is similarly possible that in both areas the main end-user of the iron was the Roman state, exploiting the metal resources in its newly reorganized territories. ⁴⁷ London’s role in the administration of mines is evidenced by an iron punch, marked M.P.BR for stamping ingots of soft metal, such as gold, as property of the mines of the province of Britain (metalla provinciae Britanniae). ⁴⁸ There is some evidence for the further intensification of iron extraction in the High Weald into the early second century, coinciding with Hadrianic reform. ⁴⁹

Were there imperial estates?

The issue of whether or not the Weald and its ironworking sites formed part of an imperial estate has divided opinion. Martin Millett correctly argues that we cannot identify such estates from the evidence of settlement archaeology. ⁵⁰ Imperial estates look pretty much like land owned by others, and the absence of villas from any given landscape indicates only that the wealth obtained from the land wasn’t being invested in this type of social display. This increases the odds that those who profited from the land resided elsewhere, but even this is not certain: architectural display is only one way in which the rich might represent social rank. On the other hand, Cleere’s argument remains a plausible speculation. The Roman state had a direct interest in controlling the mineral resources on which it relied, and important quarries and mines were often publicly owned or brought within the imperial patrimonium. ⁵¹ In principle, imperial land was no different from the estates of other private individuals. It was the Emperor’s personal property run by his slaves and freedmen, the familia Caesaris, or rented by them to private contractors (conductors). Any such agents and middlemen would ultimately have served the London-based procurator. Imperially owned quarries might also draw on the support of the provincial administration for the supply of provisions and animals, and on military specialists and aid for their management. ⁵² It can be argued that the building of metalled roads across the western Weald, and the involvement of the Classis Britannica in estate production in the south-eastern Weald, witness support of this nature.

There has been much speculation about the extent of imperial holdings elsewhere in southern Britain, notably in discussion about landscapes in the Fens and Thames estuary. ⁵³ Like the Weald, these marginal territories saw intensified exploitation in the Roman period, but little evidence of having been managed from villa estates within the orbit of self-governing civitates. Here too, there is no way of establishing who owned the land at any given point in time. The fact that we cannot securely identify specific imperial estates doesn’t mean that they weren’t ubiquitously present. The imperial government controlled a considerable part of productive land throughout the empire, such that by the fifth century the emperor is thought to have owned up to 18 per cent of the total area of two North African provinces involved in the annona supply of grain to Rome. ⁵⁴ The conquest of Britain, which was very much the emperor’s project, created golden opportunities for the expansion of his patrimonium. The property of those who fought against Rome fell forfeit as spoils of conquest, bringing extensive tracts of land under procuratorial control. ⁵⁵ Pre-conquest royal holdings may well have concentrated in strategically important borderlands flanking the Thames, where territorial annexation may have disproportionately rewarded victorious kings such as Cunobelin. Further confiscations took place when Rome assimilated allied kingdoms, as when Icenian nobles were stripped of family estates on the death of their king Prasutagus. ⁵⁶ Even more would have taken place in the suppression of the Boudican revolt, whilst the imperial patrimonium routinely incorporated property seized from convicted criminals, the intestate and tax defaulters. The conquest of Britain and disproportionate presence of the forces of imperial power are certain to have brought many lands under the Procurator’s control. As Mike Fulford and Jane Timby have asked: how much of the pre-Flavian province would not have fallen into the emperor’s estate? ⁵⁷

Much of the territory seized by Rome would have been given away. Land rewarded veterans and supported colonial settlement, and could be given to powerful allies. It could also be donated to cities as public land, and endowed to underwrite the costs of supporting temples, baths, and games. Other estates would have been sold for profit where buyers could be found. Much, however, is likely to have been retained for administration by the procurator from his offices in London. Imperial property provided income to support imperial expenditure. The procurator was responsible not only for raising taxes and revenues from imperial lands, but also for the payment and supply of the troops. ⁵⁸ The army needed grain, livestock, leather, salted goods, timber, and iron. Since the procurator was responsible for ensuring that these goods reached the army, it made sense to retain control of the estates that produced them. ⁵⁹ The acquisition and development of imperial estates are bound to have featured importantly in an economy aimed at finding local sources of goods to reduce dependency on long, exposed and costly supply lines. Flavian investment in the extraction industries and regional infrastructure can be seen as a strategic response to such needs.

The role of London

In a sense it matters not so much who owned the land where resources were found, so long as the procurator in London commanded the materials he needed. There were different ways in which this might have been achieved. The ragstone quarries and granaries of western Kent were situated within a landscape dominated by villas, and those who owned these villas are the likely owners of these natural resources in like fashion to the ownership of the wood at Verlucionium. They would have been at liberty to seek the best market and sell for cash or credit, even if some portion was diverted to settle tax and other obligations. Produce may have been sold at the ‘farm gate’ to agents with the established business contacts for its onward sale and equipped for its transport and storage. In this scenario, the procurator would simply have been the most highly valued city-based customer, securing privileged access to goods through the favoured trading relationship with middlemen that came with his powerful office.

Different arrangements would have applied on imperially owned estates. Epigraphic sources describe sharecropping on such land in North Africa during the early empire. ⁶⁰ Tenant farmers (coloni) paid rent in kind, normally set at a third of their crop, to middlemen (conductores) who held 5-year leases awarded by the imperial administration. The treasury took its rent from these conductores, probably also in kind. Similar leaseholding arrangements extended to other spheres of economic activity, including tax collection. ⁶¹ This is an attractive model to apply to the countryside around London. Sharecropping would have been an effective way of raising rent from farmers poorly integrated with a cash economy, in a countryside that was never fully monetized. ⁶² It made no demands of cash liquidity whilst reducing the administration’s vulnerability to fluctuations in supply. The emperor wasn’t the only landowner to take sharecropping rent from tenant farmers. Pliny did the same, as described in his letters, although cash rents are more widely attested. ⁶³

It was also open to the procurator to control estates within his authority using officials drawn from the imperial household or other branches of the military administration. The presence of such agents is implied by military-style buildings found at the ironworking site in the Weald at Bardown, and in the administrative compound at Mucking in the Thames estuary. ⁶⁴ In considering evidence for the grain supply of Rome, Geoffrey Rickman suggested that the late first century ad may have been a moment of transition, involving a shift from drawing supply from pubic land farmed by publicani, to a greater reliance on imperial estates run by imperial officials. ⁶⁵ Might the early Flavian changes to the landscape of control in and around London witness such a transition?

In Britain, a managed system of rent and taxes is likely to have replaced earlier reliance on unsustainable requisitioning by the occupying forces, underwritten by extended supply lines from the continent. The weight of these initial exactions is described in a statement attributed to the Briton Calgacus by Tacitus, which translates as ‘our goods and chattels go for tribute; our lands and harvests in requisitions for grain’. ⁶⁶ Tacitus describes Agricolan reforms aimed at alleviating the burdens of corn levies and taxes (frumenti et tributorum) by distributing the burden more fairly and cutting out abuses introduced by tax collectors. ⁶⁷

Throughout much of the empire local taxes were raised by members of the civitates town council or ordo, using mechanisms similar to those they employed as landowners in calling in their rents. ⁶⁸ It is open to question how extensively this model applied around London, given our doubts over the extent to which political power in London rested with local elite society. It is instead possible that direct taxes (tributum soli and tributum capitis) were collected by tax-farmers and imperial fiscal officials. ⁶⁹ It would have made sense to raise taxes in kind in newly acquired territories lacking monetized economies. ⁷⁰ There are documented instances of grain from imperial estates being supplied to soldiers, but as Keith Hopkins observed, we don’t know how much of the grain that fed soldiers had been raised as tax in kind, and how much had been purchased at fixed prices or on the open market. ⁷¹ Imperial policy tolerated taxation in kind in grain-producing provinces as a means of underwriting food supplies, but coin revenues were more valuable. Tax collectors were under pressure to convert rural produce into bullion—the stores of wealth that could underwrite wages and pay for capital projects—by selling surplus for money. ⁷² The businesses of the middlemen handling these agricultural surpluses—whether obtained through purchase, tax, or rent—were necessarily structured through town. ⁷³ This process of converting local perishable surplus into bullion/gold relied heavily on urban markets and manufacture, stimulating industrial growth in cities such as London.

The roadside and riverside settlements found around London established places where rents and taxes could be assessed and collected, serving as points of contact between London’s cash-based economy and subsistence farming communities in its hinterland. Produce could have been brought to these places by tax-paying communities, leaving the state responsible for onward transport using private contractors or soldiers. ⁷⁴ Such supplies may have made an important contribution to the grain supplies needed by the army and state. Paul Erdkamp has argued that taxation in kind, supervised by provincial administrations, was more important than market supply. ⁷⁵ The army was sometimes directly involved: an exchange of letters between Pliny when serving as governor of Bithynia and the emperor Trajan described the use of detachment of soldiers to assist the procurator in grain collection. ⁷⁶ The assertion that most grain was moved by the state rather than private trade can, however, be questioned. ⁷⁷

The complex and varied systems of rent and tax collection used in the Roman world could thoroughly distance landowners from those who farmed their land, involving a hierarchy of intermediary agents such as administrators, leaseholders, and bailiffs. ⁷⁸ This may have acted as a break on the development of rural markets amongst producer communities, whilst stimulating their development within towns. This may go some way towards explaining why many of the farming settlements near London remained comparatively unaffected, at least in terms of the material culture on display, by the presence of the city. Since our main concern is with London, in its role as a hub for the exploitation of regional resources, this chapter doesn’t explore the full range of evidence for farming in the Thames basin. It is possible that some communities retained traditional rights over the land that they farmed, little touched by the need to render tax and rent to London. It is evident, however, that the Roman conquest lead to the reorganization and intensification of regional production and that surpluses were obtained through the settlement and transportation network that focused on London. The early Flavian enhancement of this infrastructure of control is the likely product of official policy, perhaps connected to military supply and the annona.