‘The food supply of the Capital’, in: The Cambridge Companion to Ancient Rome, ed. Paul Erdkamp, Cambridge: Cambridge University Press 2013, pp. 262-277

15: The food supply of the capital Paul Erdkamp A popular modern view sees the population of the Roman capital as unemployed, degenerate and dependent on imperial largess. Few phrases from the ancient literature are so often quoted as ‘bread and games’, Juvenal’s brief expression of the corruption and corruptibility of the urban masses (10.81). In his view the Roman populace had sold to the emperors the rights and liberties they had had during the Republic. While the political dimension of Rome’s food supply is beyond doubt, one may wonder whether the grain dole is adequately explained as a bribe. The feeding of Rome, a city of nearly a million inhabitants, tested the logistical, economic and organizational abilities of the Roman world to their limits. This is reflected in every aspect of the city’s food supply: the urban diet and its shortcomings, the weaknesses of the urban food supply, the grain dole, and the state’s involvement in the acquisition and distribution of food. The urban diet The diet of all people in the Roman world consisted largely of cereals: durum wheat for the better-off people and the inhabitants of cities, emmer wheat, barley or millet for the less fortunate. However, the diet of the rural dwellers, the majority of whom were involved in farming and could gather food from the wild, was more varied than that of the urban populace. The predominance of cereals in urban consumption was partly due to logistical considerations. High transport costs favoured foodstuffs with a high nutritional value and limited bulk. Moreover, grain is less perishable than many other foodstuffs, which means that it can be transported and stored with relatively little loss. On the other hand, it was owing to the city’s status as political centre that in imperial times even the common people in Rome ate durum wheat instead of emmer, barley or millet. In contrast to the ‘lesser’ cereals, durum wheat 262 The food supply of the capital was suitable for baking into leavened bread. Hence, if the populace is said to have been corrupted by the ‘bread and circuses’ offered to them by the emperors, it does not mean merely that they could fill their bellies daily, but that they could fill them with wheaten bread, a luxury for most subjects of imperial Rome. An estimation of the amount of grain eaten on average by individuals in Rome depends on many unknown variables. While we can make reasonable assumptions about the size and stature of ancient Romans, it is more difficult to estimate their energy requirements, since we have little idea of how much physical labour the ‘average’ person living in Rome had to undertake. Even more importantly, while we can say that their diet was dominated by wheat, we cannot establish with any precision the role of other items in meeting their nutritional requirements, and we cannot express the nutritional status of people in Rome in figures. What we can say is that on average the annual consumption of wheat must have been in the range of 200 kg for an adult. Estimates in the range of 250 kg would either exclude the consumption of other foodstuffs or imply an implausibly high intake of calories, while much lower estimates would either have the people starve or underrate the dominance of cereals in the urban diet. A total consumption in the city of Rome of more than 150,000 tonnes of grain must be understood as a very rough estimate, since it is hampered by our inadequate understanding of the age structure, sex ratio and size of the city’s population. The main items of food, besides grain, were legumes, olive oil, wine and meat.1 We may be sure that these were normal items of the diet of the common people in Rome, but it must remain unclear how regular and in what quantities they were eaten. Much of the evidence on diet in the ancient world does not pertain to the population of large cities. The rations of slaves or soldiers are not relevant as far as the population of big cities is concerned. However, a number of considerations point to a diet dominated by grain and bread. To begin with, the number of instances where authors refer to the supply of grain is vastly greater than similar references to legumes, olive oil, wine or meat. The sources, ranging from Livy’s first decade on early Rome to Pliny’s eulogy of the emperor Trajan, show an almost exclusive interest in the harvest, supply and price of grain. Second, Gaius Gracchus in 123 bc started the regular distribution to the urban populace of cheap grain, and for centuries grain remained the only item that was 1 On the consumption of legumes, Corbier 1999. On meat, Corbier 1989, MacKinnon 2004. 263 Paul Erdkamp distributed in such a way. It was only from Severan times onwards that olive oil was regularly distributed, while meat and wine were added even later. Third, from the first century bc onwards, people in Rome were aroused to violent demonstrations by the high price of grain, but similar occurrences due to the high price of wine are rare and only mentioned for the fourth century ad. In short, everything indicates that the urban populace fulfilled their nutritional needs largely by the consumption of cereals.2 An overview of the diet of the people living in Rome would not be complete if we were to pay attention only to basic foodstuffs and ignore the luxurious and exotic items that were eaten in Rome too – albeit not by everyone. For example, pepper was imported from southern India by way of Egypt, a trade route that flourished from Hellenistic times until the Portuguese arrival in the Indian Ocean more than 1,500 years later. In ad 92, Domitian ordered horrea piperataria to be built. It is amazing that pepper has become so very popular, Pliny the Elder (Historia Naturalis 12.29) writes, in view of the fact that bitterness is its main characteristic. Its popularity is reflected in the recipes of Apicius, nearly all of which include pepper. Peppers have even been found in military camps along the borders. Pliny the Elder shows a knowledge of various tropical fruits, such as bananas, but supply lines were too long to allow the importation of such perishable goods. The point to be made is that, despite the general emphasis on basic staple foods in ancient sources and modern literature, the well-to-do in Rome did not regard it as depraved to eat fish, fish sauce, mussels, fowl, eggs, mushrooms, figs and dates, melons, pears and apricots.3 The dole In 123 bc the Roman popular assembly accepted Gaius Gracchus’ proposal for a grain law, according to which male citizens of Rome were given the right to buy a monthly ration of grain at a fixed and low price (6 1/3 asses). Gaius Gracchus introduced the grain dole as part of a series of measures taken to curry favour with the voters in Rome, but we should not see it solely as a bribe. Rome had grown tremendously in the previous century, and the food supply of such a large city, gradually 2 3 Cf. Jongman 2007, 603–4, who, however, does not distinguish between urban and rural consumers. On luxurious and exotic foods, see e.g. Dalby 2002. 264 The food supply of the capital outgrowing the productive capacity of its direct hinterland, became increasingly vulnerable. A measure that would create some degree of stability in the grain market was not out of place. Moreover, and this might seem paradoxical, Gracchus may have intended to de-politicize the handing out of grain. The past decades had seen the increasing tendency of magistrates to employ public grain on their own initiative for distribution among the Roman populace, but such ad hoc measures by individual magistrates did not solve the capital’s problems, while adding to those of the tax-paying provinces.4 The sources demonstrate that the grain dole remained a political issue until the end of the Republic, but the fact that we do not even know for sure what amount was stipulated by Gracchus’ grain law may serve as an indication of the fragmentary and inadequate nature of our evidence.5 The lex Terentia Cassia from 73 bc specified the monthly distribution of five modii, and it is likely that this was the amount from the start. Five modii (about 33 kg) is more than sufficient for an adult male. Such an amount might have fed two adult people, the more so if one of them was a woman. Recipients of the grain dole were male citizens from either 11 or 14 years old. In view of the much higher average age of first marriage for men, many recipients did not have a family to support. At first the beneficiaries seem to have included all male citizens of sufficient age living in Rome. There is no mention of any further restrictions. However, with regard to the lex Terentia Cassia of 73 bc, Cicero (2 In Verrem 3.72) stated that 33,000 medimnoi would be sufficient for the monthly distribution of grain to the plebs, which implies about 40,000 beneficiaries. The total number of male citizens from age 11 or 14 onwards living in Rome must have been much higher. Cato the Younger is said to have widened the group of recipients in 62 bc to include ‘the poor and landless plebs’ (Plutarch, Cato Minor 26.1), which confirms that some sort of restriction was imposed previously. In 58 bc Clodius abolished the price, making the rations a free gift. In the 40s, the number of recipients appears to have risen to 320,000, inducing Julius Caesar to bring it down to 150,000. He did this in part by conducting a census district for district, employing the services of the domini insularum, i.e. the heads of the apartment blocks that housed many of the urban poor. Augustus, following Caesar’s example, conducted a similar census and established the number of recipients at 4 5 Erdkamp 2000. See also Garnsey and Rathbone 1985. In particular, Rickman 1980a, ch. 7. 265 Paul Erdkamp about 200,000.6 We do not have similar numbers for imperial times; we can only speculate that the number of recipients did not change significantly. The frumentationes remained a privilege in later times. The recipients were entered on a list that was regularly revised.7 In the first century ad, the members of the Praetorian Guard, the urban cohorts and the vigiles were added to the list. Trajan included 5,000 children, a gesture that was repeated by some of his followers. The epitaphs of several young children tell us that they had been entitled to receive public grain, which shows that being on the list provided a certain status. Handing out grain to several hundred thousand people each month required a complex organization, spreading the actual distribution over various places and/or days per month. An inscription from the first century ad tells us that a certain individual received grain at entrance 42 on day 14 at the Porticus Minucia. Apparently there was now one distribution centre, which handed out grain to recipients allocated to specific days and entrances. The importance of the Porticus Minucia is reflected in the title of certain officials attached to the city’s grain supply: procurator Augusti ad Minuciam and praefectus Miniciae. Furthermore, a curator aquarum et Miniciae is first attested for the reign of Septimius Severus.8 In the third century the handing out of grain was replaced by that of bread, which caused a major revision of the system, which, in turn, explains the curator’s responsibility for both the aqueducts and the food supply. The emperor Aurelian is said to have handed out bread to the urban populace (Historia Augusta, Aurelian 35.1), which has often been interpreted as an innovation. However, there is good reason to date this change earlier. The distribution of bread implied a different organization, because grain could be stored for some time and thus handed out monthly, while bread required a more or less daily distribution. Moreover, the state had to take responsibility for the milling of the grain. This ties very well with the discovery of a largescale watermill in Rome that was powered by the aqua Traiana and that was constructed at some time during the third century. Its construction gave the aqueduct an entirely new purpose, and therefore the watermill must have been built on imperial orders. In view of the emergence of the curator aquarum et Miniciae under Septimius Severus, it is most likely that this development dates to his reign.9 6 8 Rickman 1980a, 175–82. Rickman 1980a, 192–7. 7 9 266 Rickman 1980a, 188–91. Bell III 1994. The food supply of the capital Septimius Severus changed the dole in another momentous way by adding a daily allowance of olive oil to that of bread. Already his predecessors in the second century ad occasionally distributed olive oil, but it was left to Severus to make this a regular gift (Hist. Aug., Septimius Severus 18.3). Alexander Severus is said to have restored the regular handing out of oil after it had been abolished under his predecessor (Hist. Aug., Alexander Severus 22.1). Finally, the emperor Aurelian is credited with adding the distribution of wine and pork to those of bread and olive oil (Hist. Aug., Aurelian 35.2 and 48.1). Vulnerability The grain dole did not feed the city of Rome. Not everyone was entitled to it, and until the reign of Septimius Severus the dole was limited to grain. The price of food was still a major concern for the capital’s populace: accustomed to buy their food day by day, Tacitus (Historiae 4.38) writes, the common people have no interest in public affairs save the grain supply. They looked to the emperor to safeguard their interests: when Nero intended to leave the city, the Roman populace feared supply problems in his absence (Tacitus, Annales 15.36.4). Several elements combined to make food prices such a pressing concern for the urban masses. First, their buying power was so low that most of their income was spent on food. Early modern figures on the consumption pattern indicate that the cost of daily sustenance used up about half or three-quarters of the income of the common people in European cities. Second, the income of many wage-earners was unstable and may have been threatened by seasonal fluctuations in employment, which were caused, for instance, by the decrease during wintertime of shipping and cargo handling or by a temporary slump in building activity. Third, when the price of grain rose, urban consumers had little recourse to alternative foodstuffs. It is clear that, if the daily sustenance consumed half of the income of an average household, the consequences of a prolonged doubling of prices were grave. Even if outright starvation was limited to beggars and vagabonds, the most fortunate among the common people had to eat into their reserves, while the less fortunate were forced to accept more frugal living conditions than they were accustomed to. Hence, price stability – or, at least, a limitation of price volatility – was something to be desired. Famine did not occur in Rome as long as its authorities ruled overseas possessions. Not until the fifth and sixth centuries, when the 267 Paul Erdkamp empire had long been lost and Italy turned into a battle zone, did the Black Horse of the Apocalypse make its appearance in Rome. However, individuals in Rome may nevertheless have died of hunger and malnourishment. The story of the beggar shooing away the dogs that are awaiting his death in order to feed on his body illustrates that abject poverty was not absent from Rome, despite the empire’s power and Rome’s status as its capital. Lack of work or the inability to work, lack of food and increasing vulnerability to disease must have gone hand in hand for many among the destitute. Disease and disability surely made it impossible for some to acquire enough food to make it through the days. Starvation, in other words, was a cause of death, albeit operating in tandem with other causes. The dominance of cereals in the urban diet must also have caused some degree of malnourishment among those who ate little else. Malnourishment is indeed indicated by ancient skeletal evidence, although it is not always possible to distinguish between the effects of a one-sided diet and those of diseases that hampered the body’s intake of necessary nutrients and minerals. Being unaware of the condition, medical writers could not recognize its symptoms, as seems to be the case in the following passage, in which the second-century doctor Soranus (Gynaikeia 2.43-4) warns about the dangers of having babies starting to walk too early: ‘If moreover it [the infant] is too prone to stand up, and desirous of walking, the legs may become distorted in the region of the thighs. This is observed to happen particularly in Rome.’ Soranus continues by rejecting various causes that have been proposed, blaming instead insufficient care by Roman mothers. As Peter Garnsey has pointed out, the passage probably points to the occurrence of rickets in Rome, a condition that is caused by vitamin D deficiency, which, in turn, may be related to the predominance of cereals in the diet.10 Most people in Rome did not have to fear actual starvation, but that did not prevent many inhabitants of Rome making their disapproval clear when the price of grain rose. Food riots in Rome were directly related to the dole, first of grain, later wine and pork. From the first century bc onwards, the sources mention grain riots aimed at those people whom the rioters thought were responsible for the situation, which meant throwing stones at the consuls in the late Republic, at Octavian and Mark Antony in the time of the second triumvirate, throwing pieces of bread at Claudius during his reign, and demanding 10 Garnsey 1999, 47–8. 268 The food supply of the capital the death of Cleander, an imperial favourite, from the emperor Commodus. In the fourth century the people responded to the high price of wine with violent protests, burning down the house of a leading magistrate. In one case we are told of demonstrations in Rome because of the high price of beef and pork (Hist. Aug., Alex. Sev. 22.7). Although the protests are dated to the reign of Alexander Severus, the passage may actually reflect the situation in Late Antiquity, when it was written. Just as in the case of wine, the regular distribution of pork altered the expectations of the Roman populace. The state had made itself responsible for an adequate food supply, and thus the people addressed the state’s representatives when reality did not meet expectations. Moreover, riots were not caused by hunger – genuinely starving people do not protest – but by the rioters’ perception of the violation of their rights. In the face of the possibility of a food riot by a violent mob, it was often opportune to live up to one’s creed and to make sure that market supply was plentiful and the price ‘just’. As the first-century bc historian Dionysius of Halicarnassus (Antiquitates Romanae 7.20) quite anachronistically wrote about fifth-century bc Rome: ‘The consuls . . . took great care to supply the city plentifully with both grain and other provisions, believing that the harmony of the masses depended on their well-being in this respect.’11 Supply and transportation Consumers in Rome ate foodstuffs from all over the Roman empire and even beyond. Pepper and other spices were imported from India and South East Asia. However, it is clear that most of the staple foods that the urban populace consumed either came from the direct hinterland of the city or from regions that offered relatively easy access to the capital. Logistical considerations determined that most of the meat eaten in Rome came from the Italian peninsula, including Cisalpine Gaul. However, Rome had become so gigantic that the food requirements of its nearly 1 million inhabitants exceeded the productive capacity of Italy. One should not forget that Rome’s hinterland counted numerous smaller towns and cities whose inhabitants had to be fed as well. In fact, Rome was only able to grow so very large because the Roman state had sufficient power and organizational ability to ensure an adequate supply of at least basic foodstuffs. Already at the end of 11 In more detail, Erdkamp 2002. 269 Paul Erdkamp the third century bc occasional shipments of grain from Sicily and Sardinia appear in the sources. Although these islands continued to play an important role in the capital’s grain supply, they were soon surpassed by Africa. Under Augustus Egypt became Rome’s largest supplier of grain, which it remained until the founding of Constantinople. Pliny the Younger observed that ‘it was generally believed that Rome could only be fed and maintained with Egyptian aid’ (Panegyricus 31). Seneca tells us that the arrival of the grain ships from Egypt in Puteoli was a cause for celebration (Epistulae 77). Africa was the second largest supplier of grain, and it also played an important role in the supply of olive oil. Most of the olive oil, however, came from southern Spain, as the Monte Testaccio – a hill near Rome consisting of millions of sherds of discarded amphorae used to transport olive oil – demonstrates. However, the produce of Italy should not be dismissed as insignificant. Many among the rich or just well-to-do families living in Rome owned land in Italy. For example, Pliny the Younger, by no means an extremely rich senator, owned property in Tuscany and Cisalpine Gaul, while Maecenas supported the poet Horace by giving him a farm in Sabinum. The urban households of the rich and well-to-do, which varied from a few relatives and slaves to palatial workforces consisting of hundreds of slaves, undoubtedly ate grain, olive oil, vegetables, legumes, meat, cheese, honey and fruit from the produce of their estates. All of the food consumed in Rome had to be transported some distance over land, if only from the river docks to the depots and from the depots to the consumers. Mosaics showing the unloading of ships demonstrate that the grain and olive oil were carried sack by sack and amphora by amphora, which is supported by the fact that the storage rooms of horrea were only accessible on foot. The movement of food offered employment for thousands of free and servile labourers. However, the food supply of Rome was only manageable if the largest part of the staple food was transported most of the way by river and sea. Rome was exceptional as a Mediterranean metropolis in the sense that it was not located downriver from fertile lands that were able to fulfil its needs, as Alexandria, Carthage and Antioch were. The upstream transportation of goods on the Tiber from the harbour at Ostia to Rome was undertaken by barges that were towed by teams consisting of either men or oxen.12 12 See Procopius, De bello Gothico 5.26.9–11 for a description of ox-drawn tow-boats in the sixth century ad. In general, see Chapter 13. 270 The food supply of the capital Ships of all sizes were involved in the provisioning of the Roman capital, but the ships bringing grain and olive oil from Egypt, Africa and Spain tended to be larger than average. The emperor Claudius offered privileges to owners of ships of at least 10,000 modii (approx. 70 tonnes) burden, while later rulings specified either one ship of 50,000 modii or five ships of 10,000 modii. Since these rulings were intended to stimulate the involvement of as many ships as possible, 70 tonnes was below average. Small freighters offered advantages to shipowners and merchants who bought and sold a variety of goods at various markets: small ships could enter shallow harbours, and it was less difficult and costly to fill the hold of small ships with goods that one hoped to sell profitably. In contrast, the volume and constancy necessary for the provisioning of Rome stimulated the use of sizeable freighters. The Apostle Paul was brought as a prisoner to Rome on such a grain ship, and his experiences, described in detail in Acts of the Apostles (27–28), graphically illustrate the physical difficulties of shipping grain from Egypt. Overseas voyages were regarded as too dangerous from mid-November to mid-March. While trying to find a safe harbour for the coming winter, Paul’s ship was blown off course and wrecked on a small island. The ship’s cargo was lost. Paul’s guards finally brought him to another Alexandrian grain ship wintering in a nearby harbour. Owing to the distance and the winter break in long-distance shipping, freighters on the Alexandria–Italy route managed no more than one return trip each year. The largest single flow of food towards Rome – the grain supply from Egypt – was a time-consuming and risky affair, requiring the continuous service of hundreds of ships. Ships from Africa and Spain managed several return trips and hence the transportation of olive oil and grain from these provinces was less time-consuming and risky. Even so, feeding 1 million people was a huge logistical challenge that required every effort of the Roman state. The question of how the authorities managed the transportation of the city’s staple foods is closely related to the question of how they acquired the necessary grain, olive oil and other goods. Opinion on this matter diverges.13 Scholars like G. Rickman assume that the authorities bought almost all the grain they shipped to Rome on the open market. He argued that the authorities required the service of a large number of merchants to supply the urban market with adequate amounts of 13 For a more detailed analysis, see e.g. Herz 1988, Sirks 1991, Höbenreich 1997, and Erdkamp 2005. 271 Paul Erdkamp grain.14 One may point to people like M. Caerellius Iazymis, who is called a codicarius, item mercator frumentarius (boat-owner and grain trader) on an inscription from Ostia. However, while we cannot deny the involvement of trade, in my view the emphasis should lie with the state, not with trade: most of the grain supplied to Rome was state-owned, and hence the authorities needed the services of shipowners who transported grain rather than that of grain merchants who supplied it. It is difficult to say when this system arose, but this is the picture emerging in the sources from the first and second century ad. Clear evidence first pertains to the reign of Claudius. Apart from building a new harbour at Ostia, he offered favourable conditions and privileges to the businessmen (negotiatores) who built ships ‘of a capacity of not less than 10,000 modii of grain, and if that ship, or any other in its place, carries grain to Rome for six years’.15 Most importantly, the authorities undertook to compensate the businessmen for any losses they incurred while serving the city’s grain supply. The text does not clearly specify who owned the grain that was to be transported on the ships involved. However, Claudius was evidently interested in the services of shipowners, not traders who did not own ships. Moreover, we know that many shipowners had contracts for transportation with the annona, the department in charge of the grain supply of Rome. Claudius limited his privileges to those shipowners who held contracts with the annona for at least six years. Similar rulings, dating to the reigns of Hadrian and Marcus Aurelius, offer similar conditions to shipowners who serve the annona.16 In short, the authorities responsible for the grain supply of Rome were eager to stimulate the cooperation of many shipowners, but there is no evidence for a similar policy towards merchants dealing in grain. Many businessmen may have been involved in both lines of business, but that does not alter the main point: the authorities, not having any freighters themselves, solicited the services of shipowners who would transport grain to Rome. The authorities did not need the services of merchants supplying grain, because under normal circumstances they had recourse to sufficient grain for the state’s overall needs. Rome’s main source of grain consisted of the taxes-in-kind that were levied in almost all provinces. Already in the Republic, Rome had introduced levies of a fixed percentage of the harvest (a tenth in Sicily and Sardinia, a twentieth in Spain) and the sources make clear that taxes-in-kind were normal in 14 16 Rickman 1980a, 72. 15 Gaius, Institutiones 1.32c. Also Suetonius, Claudius 18.2. Callistratus, Digesta 50.6.65.5, 6; Scaevola, Dig. 50.5.3. 272 The food supply of the capital subsequent provinces. Most importantly, farmers in Egypt paid taxes in grain and other crops. In Egypt taxes were fixed at a certain amount per unit of land, not a percentage of the harvest, which is undoubtedly related to the relative reliability of harvests in Egypt. Moreover, the emperors held vast imperial domains throughout the provinces, in particular in Africa and Egypt, which offered rents in kind. Even on a conservative estimate Egypt offered sufficient tax grain (some 25–30 million modii each year) to fulfil a significant part of the capital’s need.17 The sources also offer evidence of contractors who transported olive oil to Rome on behalf of the authorities, and again we may ask whether the olive oil supply of Rome reflects commercial channels that are supported by the authorities or state-dominated extraction that relied on businessmen for its distribution. One inscription, set up by scapharii (shippers of boats) in Spain in honour of Sex. Iulius Possessor, tells us that the latter, as adiutor of the annona in charge of the oil from Africa and Spain during the reign of Marcus Aurelius (ad 161-80), was responsible for the inspection of the oil, the organization of its transport and the payment of the transport charges to the shippers. The last two tasks mentioned demonstrate that the oil was already state-owned when it was transported to Rome, but the origin of the oil remains unclear. Another inscription mentions the same Possessor as adiutor of the annona connected to the depots (horrea) at Ostia and Portus, i.e. Rome’s main harbours. The state’s concern with transportation is shown by the fact that in another function Possessor was responsible for the navigability of the Guadalquivir, the most important transportation channel in Spain’s main oil-producing region. Other evidence consists of the stamps and writings (so-called tituli picti) on the amphorae in which the oil was transported. The stamps usually mention the name(s) of the owner(s) of the pottery which made the amphora, while the tituli picti contain the names of the trader or transporter, the names of the producers of the oil, and information pertaining to the amphora’s weight and content. Many of the names on the amphorae also occur in inscriptions mentioning navicularii (shippers), mercatores (traders), negotiatores (businessmen) and diffusores (exact meaning unclear) involved in the ‘oil business’. The evidence demonstrates the heavy involvement in the oil supply of Rome of people who are known to be involved in the oil trade outside of Rome. It must remain unclear, however, whether the people who traded in oil were either involved in selling oil to the state or were just 17 Erdkamp 2005, 219–37. 273 Paul Erdkamp using their ships and expertise to transport state-owned oil on the state’s behalf. It may be significant that under Septimius Severus the phrasing of the tituli picti changed: the names of private traders/transporters are replaced by the formula ‘of our imperial lords . . . ’, followed by the name of the emperor. After ad 217 the imperial names disappear in favour of a phrase referring to the imperial treasury of one of the two Spanish provinces. These changes have been interpreted in various ways, but are usually taken to indicate a more direct involvement of the state in the production and distribution of oil.18 It seems clear that to a large extent the provisioning of the capital consisted of the distribution of state-owned grain and olive oil, but the involvement of trade channels cannot be ruled out, the more so regarding olive oil, which at least in the late Republic and Principate was seen as less essential. The supply of wine and other foodstuffs seems to have depended largely on the initiative of traders who were attracted by the numbers and aggregate buying power of Rome’s consumers. Storage and distribution The state was also much involved in the distribution of foodstuffs, if only by taking care of the necessary infrastructure such as wharves, warehouses and market buildings. During the Republic the scale on which foodstuffs were supplied to the city increased tremendously, and at some time this must have led to the building of warehouses. In the early second century bc port facilities were built along the Tiber, and warehouses surely were part of this development. Gaius Gracchus is said to have built horrea as part of the grain dole that he introduced. Various warehouses are mentioned for Rome, such as the Horrea Galbae, the Horrea Lolliana and the Horrea Agrippiana. Some of the early warehouses seem to have been built by private owners, such as the Sulpicii Galbae, but already during the first century ad most of the horrea became state-owned, in the sense of having become imperial property. Colossal warehouses were not only part of imperial Rome, but also of Ostia and Portus, where many overseas imports were stored before being moved to Rome. Because the horrea were crucial for the capital’s food supply, they were built on a massive scale (thick walls, multiple stories), using the most modern techniques (ventilation, raised 18 Recently, Broekaert 2008. 274 The food supply of the capital floors). Cost was obviously no issue, and much care was given to the avoidance of spoilage or theft. Storage facilities containing dolia (large vessels for liquids) suggest the storage of olive oil and wine on a massive scale. The horrea piperataria, i.e. the ‘pepper warehouse’, indicates the need for large-scale storage for this luxury item. It was probably not only used for pepper, but also for other spices coming from beyond the boundaries of the empire. Storage space in the horrea was rented out in various types and sizes. Inscriptions mentioning the staff working there also reveal the presence of people involved in selling goods from the horrea. Although most of the goods mentioned are non-edible, it is likely that the horrea were part of the commercial distribution of many foodstuffs consumed in Rome. Unfortunately, the sources remain silent as to the exact nature of the channels through which staple foods such as grain and olive oil left the warehouses.19 At some time during the Republic a situation must have emerged in which most households in Rome relied on the service of professional bakers for their bread. Most houses in Rome must have lacked ovens, as was the case in Pompeii. ‘Rise’, Martial (Epigr. 14.223) writes, ‘the baker is already selling the boys their breakfast’. The Latin term pistor does not distinguish between the milling of grain and the baking of bread, which reflects the fact that the pistores in Rome performed both services. In view of the vulnerability of flour to spoilage, it was safest to mill the grain just before it was baked. Moreover, in contrast to later times, when milling was often powered by wind or water, Roman bakers usually employed animal labour to turn the mills. Hence, there was no point in separating the milling from the baking process. It is only in the early third century ad that we see the construction of a large-scale water-powered milling facility along the Tiber in Rome, which reflects not only the scale of consumption in Rome, but also the concern invested by the city’s authorities in ensuring a steady supply of bread.20 Various sources throughout the Mediterranean world indicate that the cost of bread was changed by varying its weight, not its price. The same is undoubtedly true of Rome: the emperor Aurelian is said to have used taxes from Egypt to increase the weight of the loaves of bread in the capital (Hist. Aug., Aurel. 47.1). The weighing of bread is depicted on the famous tomb of the wealthy baker M. Vergilius Eurysaces: large baskets filled with loaves of bread are being weighed, undoubtedly 19 Rickman 1971, 173–6. 275 20 Bell 1994. Paul Erdkamp to ensure their prescribed weight. Eurysaces was a contractor (pistor redemptor) in the urban bread supply, but it is unclear what that term exactly means. Rome numbered approximately 250 bakeries in the fourth century. Trajan organized these bakers into a collegium, which may have made it easier for the state to employ the bakers on behalf of the annona and regulate their activities. The second-century jurist Gaius (Institutiones 1.34) informs us that Trajan offered privileges to owners of pistrina who processed at least 100 modii per day and who exercised their trade in Rome for at least three years. Apparently, the authorities closely monitored the bakeries that supplied the populace with bread. They continued to do so up to the end of the western empire (and beyond), but by then they had lost their grip on the Mediterranean’s main grain-producing provinces. Shops were omnipresent in Rome, and many of them must have sold food items. Cicero (In Pisonem 67) derided one of his political opponents by stating that his bread and wine were bought at the vendor, implying that a real gentleman employed his own baker and served wine from his cellar. Nevertheless, even wealthy consumers included in their cuisine goods they bought from specialized traders, and we should not underestimate the aggregate demand from the well-to-do citizens. Columella (De re rustica 3.2.1) advised his readers that the cultivation of ‘table grapes’ was only profitable near cities where one could sell them to merchants supplying urban markets. These merchants must have catered for the needs of those middle groups who could afford them, but could (or would) not obtain grapes from their own farms. Obviously there were sellers of vegetables, fish and meat. The common people bought their pork from the butcher, Varro (Res Rusticae 2.4.8, 10) says, adding that pork was imported from as far as Gaul. Cicero (De officiis 1.152), expressing contempt for small-scale merchants, specifies fish-sellers, butchers, cooks and poulterers. An inscription from Rome (CIL 6.96830) mentions a freedwoman who sold grain and legumes (negotiatrix frumentaria et legumenaria). Alexander Severus (Hist. Aug., Alex. Sev. 33) is said to have formed corpora of wine-dealers and vegetable-sellers. Pompeii offers evidence for garlic vendors, reflecting a degree of specialization that must have existed in Rome too (CIL 4.3485. Cf. 4.202). Further reading Garnsey 1988 investigates the vulnerability of the ancient food supply to disruption and the various strategies employed to ward off hunger. 276 The food supply of the capital Erdkamp 2005 concentrates on aspects of the production, trade and distribution of grain. Rickman 1980a deals with the various aspects of the grain supply of the city of Rome. Tengström 1974 offers a study of the grain supply in late imperial Rome. A shorter introduction, offering an attempt at quantification, can be found in Aldrete and Mattingly 1999. Invaluable on diet in Rome are Garnsey 1991 and 1999. More generally, see Foxhall and Forbes 1982. Recent systematic analyses of bone data, in particular of trace elements in skeletons, add to our understanding of the diet in the ancient world. However, one should be wary of broad generalizations on the basis of small samples. General introductions are to be found in Kron 2005 and MacKinnon 2007 (see also Chapter 3). There are numerous detailed studies of the distribution of olive oil and wine in Antiquity. More general on olive oil are Mattingly 1996, Peña 1998, and Broekaert 2008. On wine, see Purcell 1985. 277