POMPEY'S APULIAN ESTATES

Abstract

Pompey owned numerous properties in Italy, but except for a few residential villas little is known of their location or economic function. However, two amphora stamps have been attributed to him, which show that he was involved in the manufacture of amphorae, and probably in the production of the wine. Four tile stamps, found in the vicinity of Gravina in Puglia and at a villa at San Gilio in the upper Bradano valley, can also be attributed to Pompey for reasons discussed in detail in this paper. Since they were found in locations close to a drove road, it is inferred that Pompey had invested in properties which could be used as pasture for transhumant sheep. The circumstances in which he acquired the estate near Gravina are discussed, and it is suggested that it fell within the territory of Silvium and is likely to have been acquired by him after the War of Spartacus in which the settlement was destroyed. It is also suggested that the estate was bought by Octavian in the sale of Pompey's properties, which was still ongoing in 44 BC.

Pompeo possedeva numerose proprietà in Italia, ma ad eccezione di alcune ville residenziali poco si sa della loro ubicazione o funzione economica. Nonostante questo, sono stati a lui riferiti due bolli di anfora che dimostrano il suo coinvolgimento nella fabbricazione di anfore e probabilmente nella produzione del vino. Anche quattro bolli su tegole, rinvenuti nei pressi di Gravina in Puglia e in una villa a San Gilio nell'alta valle del Bradano, possono essere attribuiti a Pompeo per motivi discussi in dettaglio nel presente lavoro. Poiché sono stati rinvenuti in prossimità di un tratturo per pecore, si deduce che Pompeo aveva investito in proprietà che potevano essere utilizzate come pascolo durante la transumanza delle greggi. Si discutono le circostanze in cui Pompeo acquisì la tenuta vicino a Gravina e si ipotizza che essa rientrasse nel territorio di Silvium e che sia stata probabilmente acquisita da Pompeo dopo la guerra di Spartaco, durante la quale l'insediamento fu distrutto. Si ipotizza anche che la tenuta sia stata acquistata da Ottaviano nell'ambito della vendita delle proprietà di Pompeo, ancora in corso nel 44 a.C.

It is well known that Pompey (Gnaeus Pompeius Magnus) owned numerous estates in various parts of the Empire. The details are scanty, but it is possible to get some idea of their value, at least in the case of his properties in Italy, from the figures circulating in the period that followed Caesar's assassination when the question of compensating Sextus Pompey for the confiscation of his father's estates was being debated in the Senate. They are wildly inconsistent but not irreconcilable. According to Appian (B Civ. 3.1.4) Antony proposed compensating Sextus with 50,000,000 Attic drachmas (equivalent to 200,000,000 sesterces).¹ That was not long after the Ides of March in 44 BC. But in March 43 BC when Cicero railed at Antony in the Thirteenth Philippic (11–12) for his plundering of Pompey's possessions, he claimed that the Senate had promised Sextus 700,000,000 sesterces to enable him to buy back his father's confiscated estates.² The most obvious explanation of the difference is that the Senate, in March 43, under Cicero's influence, included the value of Pompey's estates that had been acquired by Antony, whereas Antony in his dealings with Sextus Pompey in the previous year had excluded them. Appian (B Civ. 3.1.4) comments on Antony's refusal to surrender them, and Sextus may have moderated his demands at that time when he was seeking a rapprochement with Antony (Senatore, 1991: 104). In 39 BC, at the treaty of Misenum, Sextus was offered, according to Dio Cassius (48.36.5), 17.5 million drachmas (62,000,000 sesterces). Shatzman (1975: 352) argues that the figure in Dio is low because it was a compromise, and Sextus Pompey may have already received something since 44 BC. But it is also possible that the sale of Pompey's assets was still ongoing, and that 62,000,000 sesterces was the value of the estates that had not yet been sold, or that could be easily recovered. An anecdote told and retold many times has Sextus joking with Antony about the latter's possession of Pompey's house in the Carinae. It suggests that Sextus had come to accept that it would be impossible to recover his father's property in the city of Rome from Antony.³

These figures are meaningless without some standard of comparison. The most readily available is the sums that Cicero was offered in reparations for his properties in Rome and Latium when he was recalled from exile on 4 August 57 BC. In a letter to Atticus written in October 57 (Att. 4.2.5) he reports that his house at Rome had been valued at 2,000,000 sesterces, and his villas at Tusculum and Formiae at 500,000 and 250,000 respectively. Like Pompey's suburban villas they will have been at the top end of the market. Cicero grumbled that the valuation of his villas had been made very ungenerously (valde inliberaliter) by people who wanted to clip his wings, but it is likely that Pompey's estates were also undervalued when they were put up for auction and Caesar's followers were the most eager purchasers. Both sets of figures may underestimate the value of property in a normal market, but they clearly show that the total value of Pompey's properties was enormous, many times that of Cicero's.

The sources tell us a little about the location of some of his properties – generally the most prestigious which attracted eager purchasers, at least from the Caesarian faction, when they were put up for auction. The most valuable were Pompey's house and gardens in the City.⁴ Julius Caesar bought them after the battle of Pharsalus and gave them to Antony (App., B Civ. 3.3.14). He also had properties centred on villas at Alba Longa and Formiae in Latium, which were acquired by Dolabella, and in Campania, where his estate in the Ager Falernus was bought by individuals, probably brothers, with the cognomen Anser (Cic., Phil. 13.10–11). He had a villa at Alsium in Etruria (Cic., Mil. 54) and others at Naples (Cic., Att. 4.9.1), Cumae (Cic., Att. 4.10.2) and Baiae (Sen., Ep. 51.11). In Picenum he owned extensive properties which he had inherited from his father, Pompeius Strabo (Plut., Vit. Pomp. 6.1, Vell. Pat. 2.29.1).⁵ He also possessed a villa in the territory of Tarentum where Cicero visited him on his way to the East in 51 BC (Att. 5.5, 6, 7).

Cicero names Antony, Dolabella and the Anseres as purchasers of Pompey's estates in a key passage (paragraph 11) in his Thirteenth Philippic, all of them suitable targets for his invective. But, he says, there may be many others who will be driven from the properties they have acquired, who escape his memory⁶ – an obvious subterfuge to avoid offending powerful acquaintances who had bought some of Pompey's estates, as Kathryn Welch (2002: 15) has pointed out.

The sale of Pompey's properties began when Caesar came back from Alexandria in 47 (Cic., Phil. 2.64) and must have lasted for a considerable time. The size, shape and, probably, the land-use of each property had to be assessed by surveyors – metatores and decempedatores – sent out by the quaestor responsible for the sale before each estate could be advertised for auction.⁷ While this was going on, the possessions of the more die-hard of Pompey's supporters were also being put up for sale.⁸ The process was inevitably long and drawn out. In the Fourth Philippic (paragraph 9), delivered in the public assembly on 20 December 44, Cicero denounced those would-be purchasers who had been blinded by hope of plunder and had not been sated by the endless auction of confiscated goods: they thought they would lack nothing to seize so long as Antony was around. But even after Antony had left for Cisalpine Gaul, the sale of the confiscated estates continued, and when Cicero delivered his Eighth Philippic on 3 February 43, he claimed that Antony was still encouraging his supporters to hope for spoils, and that his band of robbers were marking out for themselves the best houses, gardens and estates at Tusculum and in the Alban Hills. Even rustic types (homines agrestes) were making their way to the waters (Baiae) and to Puteoli in vain hope. The process must have been interrupted in April 43 by the Mutina War and the events that followed it, and by the end of the year it was superseded by the new and much more extreme wave of proscriptions and confiscations introduced by the triumvirs after they had consolidated their power in the lex Titia passed on 27 November 43 (App., B Civ. 4.7).

Production on Pompey's estates

The most valuable of Pompey's properties must have been his house and gardens in the city of Rome.⁹ They were primarily residential, or intended for display, as probably was his suburban villa in the Alban Hills where he received Crassus and his entourage in 55 (Cic., Att. 4.11.1). But most of his other properties known from the literary sources are likely to have been productive estates centred on villas which Pompey could visit from time to time and where he could entertain privileged guests, as was the normal practice among the Roman aristocracy. But his vast wealth must imply that in addition to these prestigious properties he owned many others which were primarily productive units managed by his dependants – freedmen or slave vilici, where he had no need of luxurious buildings. They are likely to have been distributed throughout Italy and must have been acquired on a rational principle since, according to the elder Pliny (HN 18.35), he never bought land belonging to a neighbouring estate. That was no doubt to guard against crop failure in the local climatic environments typical of Roman Italy.

The literary sources for the confiscations tell us nothing about his estates as economic units of production, but a little can be learned from two amphora stamps which have been attributed to him. One is on the rim of a Lamboglia 2 type amphora fragment found in the sea off Torre Valdaliga near Civitavecchia (Desy, 1989: 47 no. 2311); the other is on the rim of a fragment of a Lamboglia 2 variant found at Suvaki on the island of Pantelleria (Manacorda, 2005; Amela Valverde, 2011). Both appear to have been made from the same die. The complete text reads CN⋅P⋅MAG which in all probability stood for Gn(aeus) P(ompeius) Mag(nus) – or more probably Gn(aei) P(ompei) Mag(ni) using the genitive of possession to indicate Pompey's ownership of the amphora and/or its contents. The stamps show that Pompey was involved in the manufacture of amphorae, and very probably also in the production of the wine that they contained. Lamboglia 2 amphorae were normally made on the Adriatic coast of Italy, but analysis of the fabric of the fragment found on Pantelleria shows that it contained volcanic inclusions characteristic of the Vesuvian region. It is likely, therefore, as Amela Valverde (2011) has argued, that Pompey was producing wine in Campania, probably on his Falernian estate, and marketing it in amphorae made in Campania by potters who were more familiar with the type of amphorae used in Picenum where he also owned extensive properties. He perhaps brought in potters from his Picene estates for that purpose. The find-spots of the stamped fragments show that the wine was being traded both to Gaul or the western Mediterranean by way of the Etruscan coast, and to North Africa or the Eastern Mediterranean by way of Pantelleria.

Pompey's landholdings in South Italy

Other evidence suggests that Pompey also owned estates in South Italy, in addition to his villa in the territory of Tarentum. There is inscriptional evidence for a number of individuals with the gentilicium Pompeius in and around Tarentum,¹⁰ some of whom may be the descendants of slaves employed on his estates and liberated by him. They are likely to include captives taken in Pompey's eastern wars, such as the old Corycian man whom Virgil envisaged seeing below the walls of Oebalia (Tarentum).¹¹ Servius suggests that he was one of the Cilician pirates captured by Pompey in the war of 67–66 BC and resettled by him on land in Greece and Calabria (i.e. the Salentine peninsula, including Tarentum).¹²

There is, however, much controversy on the nature of this settlement. According to Plutarch (Vit. Pomp. 28) more than 20,000 Cilician pirates were resettled by Pompey, some at Soli and other half-deserted cities in Cilicia, and the majority in Dyme in Achaea. He says nothing about the settlement in Calabria. It is possible, therefore, that only a remnant was settled in the vicinity of Tarentum, and that they were planted by Pompey on his own estates.

There is some evidence to suggest that Pompey was involved in reorganizing the municipal government of Tarentum. The constitution of the city must have been changed at some point in the course of the first century since the ruling magistrates imposed by the lex municipii Tarentini after the end of the Social War were quattuorviri,¹³ as they were in most other municipia created at the time, whereas those attested by an inscription of the late first century were duoviri (CIL I² 3169), more typical of a colonial foundation. Marina Silvestrini (2013: 703) has argued that two fragmentary inscriptions which appear to associate Pompey with a duovir named Cormus (a rare name of eastern origin, attested in Asia Minor) provide at least indirect evidence that the municipal government of the city was reorganized under Pompey with duoviri replacing the quattuorviri.¹⁴ E. Lippolis (2002: 160–1; 2006: 45) argued that Tarentum must have been refounded by Pompey with a new deductio of settlers (including the Cilician pirates), which would have been validated in the consulship of Crassus and Caesar in 59 as part of the legislation which ratified Pompey's general settlement of eastern affairs; and he suggested that the urban plan of the city was redesigned at this time. But there is no indication of a colonial deductio at Tarentum under Pompey in the liber Coloniarum, and others have been more cautious, calling into question the date of the remodelling of the urban plan and noting the lack of any definitive evidence to show that Pompey refounded the city as a colony.¹⁵

Moreover there are various other examples of communities in Central and South Italy which were reorganized in the late Republic under duoviri and never gained colonial status.¹⁶ Of special interest here is Aceruntia (Acerenza) in the northeast quadrant of Lucania, where an inscription found in 2011 published by Marcella Chelotti (2015) records two quattuorviri who saw to the construction of a bath building with its various components including a pool (piscina).¹⁷ Another inscription from Aceruntia (now lost) recorded duoviri Gnaeus Pompeius and Marcus Satrius who restored a piscina.¹⁸ Unless it refers to a different piscina, it must be dated after the inscription recording its construction by quattuorviri, and it shows that the constitution of the municipium was changed at some time, perhaps around the middle of the first century. The filiation and cognomina (if they had such) of the duoviri are lacking in the inscription, so there is some doubt about their kinship and social standing. Bispham (2007: 386) suggests that Pompeius may be Magnus himself, or more probably his son, or the suffect consul of 31 BC; but it is also possible that he was a freedman of Pompey's who had been set up by his patron as a landowner in the community. Whatever the case, it suggests that Pompey had a role in reconstituting the municipium which he no doubt controlled through his dependants.¹⁹

Pompey had at least one estate in Lucania attested in the literary sources. It was acquired at the auction of his properties by his former slave Draco, who had changed sides and been liberated by Caesar. Cicero (Phil. 13.12) referred to him as a dragon (draco) who had seized Pompey's Lucanian possessions and embraced the patrimony of his master as a dragon does treasure.

It has been suggested that Pompey inherited some of his South-Italian estates from his great-uncle, the poet Lucilius, who is believed to have owned large properties in Apulia and Lucania.²⁰ He came from a senatorial family, but chose not to follow a senatorial career, and remained a rich eques. An anecdote recorded by Cicero in the De oratore, if it is correctly attributed to Lucilius, implies that he owned large flocks of sheep (or herds of cattle) which he grazed on public land, exceeding the quota laid down by the Licinian-Sextian laws.²¹ In fragments of his Satires Lucilius refers to his farm manager (vilicus) and his cowherd (bubulcus) – all of which suggests that part of his wealth was invested in stock-raising and that he owned a villa or villas, as well as grazing his animals on public land. There is no precise information as to where they were located, but various references in the Fragments to Apulia, Bruttii, Sicily and Sardinia show that he was familiar with these regions and suggest that he owned estates in each of them. In particular he appears to have a good knowledge of Tarentum, which strengthens the idea that Pompey's villa there was inherited from him. If the reading Lucilius in the De oratore is valid, then we can probably infer that Pompey inherited a stake in the transhumance economy from his great-uncle.

The evidence of tile stamps

In a previous publication (Small and Small 2022: 185–6), Carola Small and I suggested that three fragments of stamped tiles found on sites in the territory of Gravina in Puglia can be taken as evidence that Pompey owned a large estate in the vicinity of Roman Silvium. The argument, however, is complex and needs to be examined in detail.

The fragments were found on two sites situated on a plateau above the left bank of the Basentello river (Fig. 1). All come from flat tiles which, to judge by their thickness and typical reddish fabric, were probably tegulae, although no trace of their raised lateral flanges is preserved on any of the pieces. All three were stamped with letters in low relief inside a rectangular frame. Such tile stamps are relatively rare on South-Italian sites, where only a tiny fraction of total tile assemblages carries stamps (Small, 2005). Those that survive generally give the name of the owner of the estate on which the tile kiln was situated, who would also have owned the kiln and the slaves who worked in the production process. Landowners frequently had kilns built to produce tiles for use on buildings inside their estates, though they no doubt sometimes sold them to other landlords or building contractors in the vicinity. Given the cost of transporting heavy items by land, tiles are unlikely to have travelled far in areas where there was no possibility of moving them by water. A stamped tile is most likely, therefore, to give the name of the owner of the estate on which it was found, though the possibility that it was brought in from another estate in the vicinity cannot be ruled out.

Fig. 1. The find-spots of Tiles nos 1 and 2. Map by Carola Small.

TILE NO. 1

This fragment was found in the remains of a Roman villa located on a spur of the plateau of San Felice, 11 km west of Gravina. It has been reported by the excavators Hans vanderLeest and Myles McCallum (McCallum and vanderLeest 2014: 125–7). The stamp (Fig. 2) measures c. 2.5 cm in height and has a preserved length of c. 8 cm, but it is broken short at both ends. The letters are of uniform depth (c. 0.3 cm) and close set, but they differ considerably in width. Most are easily read: a narrow C, occupying less than half a circle, followed by an N and an M. The legs of the M are well spread, and the V between them descends to the base line. There follows an A, tilted slightly so that the left leg is almost vertical, and the horizontal bar is set low. This is followed by another narrow C which merges with the vertical bar of the defective final letter. A comparison with Tile no. 2 (below) shows that this feature is intentional and that the letter is to be understood as a G. The final defective letter has a vertical bar joined at the top by part of another bar which slopes obliquely downwards for a short distance to the broken edge. It is likely to have been another N, in which case the entire stamp can be read CN⋅MAGN. The tile was found in a context of Phase I dated by the excavators between the mid-first century BC and the early first century AD (McCallum et al., 2011: 36). If, as seems probable, it was used in the construction of the villa, then it should belong to the beginning of this phase and it is one of a very small number of stamped tiles in South Italy that can be dated by an excavation context to the Late Republic.²²

Fig. 2. Tile no. 1. Photo courtesy of Myles McCallum.

TILE NO. 2

The second fragment (Fig. 3) is a surface find collected on the same site as Tile no. 1 by the late Antonio Florido of Gravina at some time before the beginning of the excavation (Small, 2014: 73–4 and fig. 1 (P1376)). It is broken short at the left end, but the preserved letters are similar in shape and size to those of Tile no. 1. They read PM'G⋅P. The two Ps have a vertical bar and an eye formed by a line which descends at an angle and then turns inward to form an open triangle. The M is formed like the M of Tile no. 1 with upright strokes well spread, and the V between them descending to the base line, but in this case the right bar of the M has an oblique upward-trending appendage. In this it resembles the standard abbreviation for the praenomen Manius, and it is probable that here too it stands for a combination of M and A. The G is formed like the penultimate letter on Tile no. 1, except that there is a short bar set at an angle to the bottom of the letter which links it with the vertical bar of the P (a feature seen more clearly on Tile no. 3). There is an interpunct dot between the G and the P. The preserved part of the text can therefore be read PMAG⋅P. The beginnings of another letter can be seen at the extreme left edge of the sherd.

Fig. 3. Tile no. 2. Photo by author.

TILE NO. 3

The third fragment (Fig. 4) was found on another site c. 2 km to the north, Site 704 in the inventory of the Basentello Valley Field Survey (Small, 2014: 73 and fig. 1 (P1783); Small and Small, 2022: 637 no. 2225). The stamp corresponds closely to that of Tile no. 2. It too is incomplete, but in this case it is the right end of the stamp that is missing. At the left end there is a blurred M, the extreme right edge of which is visible in Tile no. 2. The other letters correspond to those on Tile no. 2, although part of the ‘eye’ of the first ‘P’ has been lost. The final P is missing, but the angular line that linked it with the G is preserved, as is the interpunct dot between the G and the (missing) P. The preserved text of Tile no. 3 therefore reads MPMAG⋅[-.

Fig. 4. Tile no. 3. Photo by author.

Fig. 5 shows the relationship between the two stamps, with the stamp of Tile No. 2 superimposed over the corresponding part of tile No. 3 (Fig. 5a), and vice versa (Fig. 5b). The measurements are identical in the areas of overlap and the letterforms correspond closely, if allowance is made for the differing degrees of wear. The close similarity suggests that they were made from the same die-stamp, and the complete text reads MPMAG⋅P.

Fig. 5. The stamps of Tiles nos 2 and 3 superimposed on each other.

Interpretation

According to McCallum and vanderLeest (2014), Tile no. 1 is stamped ‘CN/MAG[N]…’, which, they argue, may be associated with Gnaeus Magnus, who was either the owner of the land on which the tile was produced or the purveyor of the tile-yard that produced it; and in a footnote they say that it is tempting to associate it with Pompey (Gnaeus Pompeius Magnus), who wintered his troops in Apulia at the outset of the Civil War with Caesar.²³ The temptation is indeed strong since, as Syme pointed out (1958: 172–4), it was normal practice to refer to powerful nobiles in the last phase of the Republic by their praenomen (abbreviated) and cognomen, omitting the gentilicium. Obvious examples include ‘L. Sulla’, ‘P. Scipio’, ‘C. Caesar’ and ‘M. Metellus’. Syme also adduces Pompey's older son, Gnaeus, who minted a series of denarii in Spain before the battle of Munda, which bore the legend CN⋅MAGN IMP or CN⋅MAGNVS IMP.²⁴

Numerous sling-shots found in Spain are stamped with the legend CN MAG. Most have no precise provenance and can be dated only by broad geographical considerations and by what is known of the military campaigns of the civil wars in Spain. By far the largest group comes from Andalucia in Hispania Ulterior. They can be confidently associated with the campaign of Gnaeus (Pompeius) Magnus junior against the Caesarians that ended in the battle of Munda in 45 BC;²⁵ but a small group found near Tortosa in Hispania Citerior are more likely to have been used by slingers fighting under the elder Pompey's legates Afranius and Petreius, who were charged with holding Spain against Caesar's invasion and were defeated at Ilerda in 49 BC (López Vilar, 2013: 437–9). The inscriptions on these most probably refer to Gnaeus (Pompeius) Magnus senior. They show that Pompey's son adopted his father's nomenclature, and that on mundane objects Pompey senior's name might be abbreviated to CN MAG as well as to the CN⋅P⋅MAG used on his amphorae.

It is therefore a reasonable proposition that the legend CN⋅MAGN[… stamped on Tile no. 1 from the villa at San Felice also refers to Pompey. In formal and official contexts Pompey was given his full name and titles, as on an honorific inscription from Auximum datable to 52 BC by the reference to his third consulship: [Cn(aeo) P]ompeio Cn(aei) [f(ilio)] [Ma]gno, imp(eratori), co(n)s(uli) ter[tium], [pa]trono publice.²⁶ But stamps used on sling-shots or tiles required a much briefer text. In these contexts, it suited Pompey to abandon the gentilicium, which he shared with numerous freedmen, in favour of the cognomen Magnus which emphasized his unique dignitas. According to Plutarch (Pomp. 13) he was greeted by Sulla as Magnus in 81 BC on his return from North Africa where he had defeated the last of the Marians; but he was then a very young commander, and he refrained from using the name in his own documents until 77 BC when he was sent as proconsul to Spain to pursue the war against Sertorius – by which time the cognomen had become familiar and was no longer a cause of jealousy. It is likely, therefore, that stamps which identify Pompey as Magnus should be dated after 77 BC.

The Correspondence Between Tile No. 1 and Nos 2–3

Stamps nos 1 and 2 were found on the same site, their letter forms are broadly similar (the M and G) and both include the string MAG. It is worth investigating, therefore, whether the stamp of tiles nos 2 and 3 may not also refer in some way to Pompey, in spite of the fact that in place of his praenomen CN(aeus) it has the letters MP. P can hardly represent a praenomen (Publius) since it is preceded by the M. It could, however, stand for the gentilicium, Pompeius, just as it does in the two amphora stamps referred to above, which read CN⋅P⋅MAG, and on a semilunate bronze tablet inscribed with the letters CNPMAGNVS incised with dots without interpunct marks, and with the N of the praenomen and P of the gentilicium in ligature, which was recovered from the sea off Capo Rasocolmo near Messina and which Manacorda (2005: 138 and fig. 5) suggests was a label attached to a small bust of Pompey. But the M which replaces CN on our tile stamps still needs to be accounted for. An M in this position would normally stand for the praenomen Marcus, but if that is the case here then the stamp can hardly refer to a close relative of Pompeius Magnus since the name Marcus is not attested in his branch of the gens Pompeia. M could conceivably be the abbreviated name of a slave figulus who was the property of Pompeius Magnus, though even in that case it could hardly stand for Marcus, which is an improbable name for a slave. It could, perhaps, be an abbreviation of a more typical slave name beginning with M, such as (hypothetically) Maurus, in which case the stamp might be read as M(auri) P(ompei) Mag(ni): [the tile] of Maurus [slave] of Pompeius Magnus. But although slave names are frequently found on amphorae in this period, they are usually written in a more extended form which leaves no doubt as to their meaning. We may compare the stamp SOCRAT ⋅ CRS, seen on Lamboglia 2 amphorae from Reggio Emilia and Taranto, which has been interpreted as Socrates, the slave of [M. Licinius] Crassus, Pompey's colleague in the consulship of 70 BC (Cipriano, 1994: 212; Amela Valverde, 2011: 196). But the analogy is misleading since amphora stamps are indicative of a level of organized production linked to a widespread market in which the names of the slaves who had special responsibility for the production process were given prominence,²⁷ and they have little relevance to the brick and tile industry of the Late Republic in South Italy, where tiles were still manufactured and produced locally for use in the immediate vicinity. As we have seen, tile stamps of the period generally gave the name of the dominus who owned the land on which the kiln was built, and it was not until after the beginning of the imperial period that the names of slaves began to appear on the products of South-Italian tile kilns, either on their own or alongside those of their masters.

This last assertion is impossible to prove definitively since only about a dozen tile or brick stamps with slave names are known from South Italy and most of these are surface finds or come from badly reported excavations and cannot be dated by their context. In some cases, however, the nomenclature of a stamp on a surface find implies that it is imperial in date, as in the case of a tile stamped by Hermes, slave of Tiberius Claudius Diadumenus, found at Gaudiano in the territory of modern Lavello, where there was an imperial estate (Morizio, 1990: 47–8, no. 5). Others can be dated to the imperial period by excavation contexts including a tegula stamped by Philomusus found in the Casa dei Mosaici at Grumentum, which is dated broadly to the first or second century AD (Bottini, 1997: 196 no. 5), or another with the fragmentary stamp POM͡AṚ[- from the villa at San Gilio in the upper Bradano valley, which is interpreted by Helga di Giuseppe (2008: 347) as belonging to a slave called Pomarius and dated to the first century AD. It is surely indicative that a tile used on the imperial estate centred on Botromagno was stamped with the name of Caesar Augustus, whereas those produced at Vagnari 11 km away were stamped by Gratus, slave of Caesar. These should probably be dated to the reign of Tiberius or a little later, and the change from the one type to the other is likely to reflect the increasing complexity in the production and distribution of tiles on this imperial estate.

The hypothesis that the M in the stamp MPMAG.P stands for the name of a slave is therefore unlikely, given the probable date of our stamps. An alternative explanation is therefore needed. It may be suggested that the M refers to Pompey's third wife Mucia Tertia, daughter of the pontifex maximus Quintus Mucius Scaevola and mother of Gnaeus and Sextus, who continued the civil war after their father's assassination in Egypt. Pompey married her in 80 or 79 BC and divorced her in 61 BC to marry Julia, the only child of Julius Caesar, when they formed the so-called triumvirate with Crassus. She would have been known as Mucia Pompei – Mucia (wife of) Pompey, following the practice common in this period by which women married to men of high rank identified themselves by attaching their husband's name to theirs in the genitive case. The most famous example is the inscription on the tomb of Caecilia Metella at Rome: CAECILIAE | Q⋅CRETICI⋅F | METELLAE⋅CRASSI: [the tomb] of Caecilia Metella, daughter of Quintus Creticus, (wife) of Crassus – Crassus being M. Licinius Crassus, son of Pompey's colleague in the consulships of 70 and 55 BC (CIL VI 1274 = CIL VI 31584). Such marital names are well attested in lapidary epigraphy, but they are only occasionally found on brick and tile stamps. Silvia Braito (2020: 50) in her recent detailed study of female entrepreneurs in Roman Italy lists six instances. The earliest, Calpurnia Corvini (no. 28 in her catalogue), might be identified with the wife of Marcus Valerius Messalla Corvinus, consul in 31 BC, though it could also be attributed to the wife of his namesake, consul in AD 58.²⁸

If this explanation is right, the first part of the tile stamp could be read either as M(ucia) P(ompei) Mag(ni): Mucia (wife of) Pompeius Magnus) or as M(uciae) P(ompei) Mag(ni) with Mucia's name in the genitive implying that the tile stamp was hers. The coupling of Mucia's name in abbreviated form with Pompey's on the stamp would suggest that Pompey had made her his agent or even his business partner during one or other of his long proconsulships abroad, either in Spain (77–71) or in the East for his campaign against the pirates (67) or in the Mithridatic War (66–62). We may compare the role that Cicero's wife Terentia played in managing her husband's affairs during his exile in 58–57 and again after his flight with the Pompeians to Greece in 49–48 (Treggiari, 2007: esp. 60, 86, 112).

On this assumption, the tile stamps would have to be dated before 61 when Pompey divorced Mucia, and they would provide an early example of a woman who owned or administered an estate with brick- or tile-works in the Late Republic and Early Empire. In her work on private domini in Roman brick-stamps, P. Setälä (1977: 211; 2002: 184) calculated that out of 150 known domini, 50 were women,²⁹ but that was probably an underestimate. Braito lists 175 women known from their stamped names to have been involved as owners (dominae) or managers (officinatrices) in opus doliare – principally in the production of bricks and tiles, but also of architectural terracottas, amphorae, dolia and mortaria. Most of her data belong to the first and especially second centuries AD and are drawn from the city of Rome and its environs, but there are instances from all regions of Italy except for Transpadana and Liguria, and some can be dated to the beginning of the imperial period. The best-known, perhaps, is Holconia, daughter of M. Holconius Rufus, the public priestess (sacerdos publica) of Pompeii near the end of the first century, whose ownership of brick or tile works is recorded on several pieces stamped HOLCONIAE⋅M(arci)⋅F(iliae) (CIL X 950; Castrén, 1975: 71 and 176, no. 197.5). Another is Minatia L(ucii) f(ilia), who is attested on stamps found in Venosa (Roman Venusia) and the surrounding area. She is most likely to have lived around the end of the first century BC or beginning of the first century AD (Chelotti, 2003: 65–7; 2018: 185–7), as probably did Licinia L(ucii) f(ilia) Secundilla, who is also attested on a tile stamp from the territory of Venusia (Chelotti, 2003: 68; 2018: 188). But they might be a little earlier since both were making tiles needed in the colony refounded by the triumvirs for the veterans of the battle of Philippi in 42. Titia, whose slave Phileros stamped tiles at Grumentum, is also likely to have lived in the Augustan period or shortly afterwards (Braito, 2020: 296–7, no. 139).

It might be supposed that the fact that it is impossible to point to any certain examples of female dominae in tile stamps earlier than the Augustan period suggests that there was none; but that argument is invalid because when names on tile stamps were abbreviated by omitting their grammatical endings, as they frequently were, it is normally impossible to know whether the name was male or female, as Braito has noted (2020: 45). The problem is well illustrated by a stamp found in the excavation of the Roman necropolis of Monte Carru in Alghero, Sardinia. It has been published as Fundan(ii) s(ervus)/Tarrens(is): Tarrensis, slave of Fundanius, but other evidence, discussed by P. Longu and P. Ruggeri (2019), suggests that the owner of the slave Tarrensis was none other than Fundania Gallia, the wife of the polymath M. Terentius Varro, who dedicated the Res Rusticae to her in 37. Moreover, if only the ending of a name is preserved in the masculine genitive case, the possibility cannot be excluded that the complete stamp may have had a woman's name attached to her husband's in the possessive genitive.

Various studies of the rights of women under Roman law have shown that most aristocratic women had already acquired the right to own and administer property by the end of the second century. If they were married sine manu, they theoretically remained under the legal potestas of their father or other agnatic male relative, or they were assigned to a male tutor; but although women continued to need male guardians to represent them in the law courts, in other matters the rights of male guardians were easily ignored (Gardner, 1999; Berdowski, 2007). There is therefore no reason why a female domina should not have had her name represented in tile stamps already in the middle of the first century, albeit in cryptic form.

Other possible interpretations of the M in the tile stamp might be suggested, though they are less probable. It might, for instance, stand for mancipii in the sense of ‘of the slave’, or merx meaning merchandise, item of trade, but there is no parallel for either usage in brick or tile stamps. The question cannot be decided absolutely, but enough should have been said to show that, in spite of the uncertainty of the reading, the stamp can still be associated in some way with Pompey.

There remains the problem of the final P, which is separated from the rest of the text by an interpunct dot on the stamp used in Tiles nos 2 and 3. It is impossible to know whether it also appeared on Tile no. 1 on which the right end of the stamp is missing. The letter P is used as an abbreviation for a wide range of terms in Roman inscriptions. Tom Elliott (1998) lists 3,479 instances derived from an electronic search of the volumes of L'Année épigraphique for the years 1888–1993. Clearly P could stand for a large number of different terms. Most of those given in standard lists of epigraphic abbreviations (e.g. Cagnat, 1898: 376–445; Sandys, 1927: 294–311) can be ruled out as unsuitable in this context, but a few need some comment. In Roman brick stamps of the second century AD, the letter P is frequently used as an abbreviation of praedia, but normally in the phrase ex p(raedis) (‘from the estate’) followed by the name of the estate or of its owner, so identifying the estate where the brickworks were situated. This can hardly be its meaning here since the preposition is lacking and the notional praedia are not identified.

Similarly, it is unlikely to stand for posuit (he/she placed/erected), which was frequently abbreviated to P and used in the imperial period to indicate that the dedicator of a funerary stele had erected it for a deceased person. The verb was also used to record the erection of a building, but neither usage is appropriate for a tile stamp; and other verbal forms which might be envisaged, such as praebuit (he/she provided), produxit (produced), probavit (approved for use), can be ruled out as implying a degree of involvement by the dominus/domina in the production process, which is improbable in the social context of this period.

The P could theoretically be an abbreviation of an additional name, an agnomen. Agnomina were normally aristocratic and of three kinds: adoptive, indicating the gens of the family to which the individual adopted had belonged at birth; triumphal, indicating the geographical area where he had earned his triumph; or personal, indicating some trait of his character or physique. The first of these can be ruled out. The second is worth considering since the P could conceivably stand for Ponticus, commemorating Pompey's triumph over Mithridates king of Pontus, which he celebrated in 61 BC. But it is unlikely since Pontus was only one of numerous regions listed on the placards carried in his triumphal procession,³⁰ and there is no evidence that Pompey used any triumphal term as an agnomen. It is even less likely to stand for the character trait Pius, which was used as an agnomen by Q. Caecilius Metellus, the partisan of Sulla, to emphasize his support for his exiled father, Metellus Numidicus (Vell. Pat. 2.15.3). Pompey's son Sextus also used it to proclaim his loyalty to the memory of his father and brother when he continued the war against the Caesarians after the battle of Munda.³¹ But Pompey senior had no reason to use the agnomen Pius to demonstrate his piety towards his family.

Proconsul?

A more probable explanation is that the letter refers to a public office held by Pompey and follows his name according to the normal practice in Roman titulature. It can be seen on some tile and brick stamps dated by the consules ordinarii of the year. There are notable series from Veleia, where the stamps can be dated (with gaps) between 76 and 13 BC, and from Rome where numerous bricks made for use in the city and its environs were stamped with production details including the names of the eponymous consuls between AD 110 and 164.³² At a more humble level, a series of five tiles found in a context of the first century AD in the kilns of Giancola near Brindisi are stamped with the names of two municipal magistrates, Lucius Audius and Lucius Graeceius: L(ucio) Audio L(uci) filio IIIIvir(o)/L(ucio) Graeceio L(uci) f(ilio) quinq(uennale). They may have been made for a public building commissioned under their watch (Manacorda and Pallecchi, 2012: 190–4). More rarely, a magistracy which an individual had held might be recorded on a tile for reasons of prestige rather than to indicate its date, which might be some years after his tenure of the office had ended. This must be the case with Q. Laronius, who fought under Agrippa in the Sicilian campaign against Sextus Pompey and was appointed suffect consul in 33 BC. He owned estates in the vicinity of Vibo Valentia and Croton and made numerous benefactions to his homeland, which are recorded by tiles stamped Q ⋅ LARONIVS ⋅ COS ⋅ IMP ⋅ ITER (‘Quintus Laronius consul and imperator for the second time’) (Perotti, 1974; Paoletti, 1994: 490).

If the final letter P on our tile stamps denotes a public office held by Pompey, it can only refer to his proconsulship. There was no standard abbreviation of the term proconsul: PROCOS, PROCO, PROC and PR all occur in Roman epigraphy. No doubt because of the problems of ambiguity the term was not normally abbreviated to P before the middle of the third century AD when the imperial appellation pater patriae, proconsul, which had generally been abbreviated as p. p. proc., began to be rendered as p. p. p. on military diplomas.³³ But given the propensity of the makers of tile stamps to abbreviate words, it would not be surprising if the artisan who made the die-stamp used for these tiles had already reduced proconsul to P in the time of Pompey, anticipating the development of the third century AD. The last part of the stamps we have been investigating would then be read as MAG(nus)⋅P(roconsul).

The hypothesis, however, raises the question of how proconsulships were alluded to in inscriptions of the time of Pompey – not an easy question to answer since the term was in a process of linguistic transition from being a phrase consisting of preposition and noun (pro consule, abbreviated to pro cos, usually with an interpunct dot), to being a composite noun (proconsul, abbreviated to procos). It is clearly more likely that the term would be abbreviated to a single P if it was regarded as a unitary noun at the time when our tile stamp was made. There can be no doubt that the normal form in the Late Republic and throughout the early Principate was pro ⋅ cos.³⁴ But procos is also found, although the evidence is not always clear. A particularly critical case is an inscription datable to 135 BC, which records that Sextus Atilius Saranus, son of Marcus, proconsul, established boundaries between the communities of Vicenza and Este in Cisalpine Gaul. Theodor Mommsen in the first edition of the Corpus Inscriptionum Latinarum published in 1863 transcribed the first line of the inscription as SEX ⋅ ATILIVS ⋅ M ⋅ F ⋅ SARANVS ⋅ PROCOS … (CIL I 549). In the second edition, published in 1918, Ernst Lommatzsch amended the text to read: SEX ⋅ ATILIVS ⋅ M ⋅ F ⋅ SARANVS ⋅ PRO ⋅ COS (CIL I.2 636). But it is clear from the photograph of the inscription published by Attilio Degrassi in the Inscriptiones Latinae Liberae Rei Publicae, Imagines (1965: 145 no. 203) that Mommsen was right and that there was no space between PRO and COS, let alone a space and a dot. Another early instance of PROCOS more directly relevant to Pompey can be seen on two sling bullets used by Quintus Sertorius, the last leader of the Marian faction, in resisting Pompey's army in Spain, which had the legend Q ⋅ SERTOR / PROCOS stamped on one side and PIETAS on the other (Beltrán Lloris, 1990; Keppie, 2023: 23–30 and fig. 13.1). The inscription challenged the legitimacy of Pompey's unprecedented proconsulship given to him by the Senate to take over the command of the campaign.

There are other more doubtful instances,³⁵ but these examples, few though they are, show that the unitary word proco(n)s(ul) was sometimes used in inscriptions of the Late Republic. There is no doubt that pro consule was more common. It was the term used (abbreviated to PRO ⋅ COS) by Pompey himself on coins minted for his campaign against Julius Caesar in 49 BC (Crawford, 1974: 463, nos 446, 447), but as we have seen there was no consistency in these matters, and it remains a reasonable conjecture that the letter P on our tile stamps alludes to Pompey as proconsul.

It is easy to see why Pompey (or more probably his loyal agents) might wish to record his proconsulship, even on mundane objects that would be seen and understood only by a limited number of individuals involved in the construction works on his estates. For most of his career the proconsulship was the basis of Pompey's power, and the various enactments which gave him proconsular imperium violated the traditions of the Republic on an increasing scale. In 77 BC he was made proconsul by a decree of the Senate proposed by Lucius Marcius Philippus to enable him to take over the war against Sertorius in Spain. There was no constitutional precedent for this appointment since Pompey had not yet held any of the preliminary magistracies, and indeed had not yet entered the Senate. He was still holding the proconsulship in 71, when he was recalled to Italy by the Senate to join Crassus in the war against Spartacus, and he did not relinquish it until the end of the year when he gave it up in order to hold the consulship of 70, jointly with Crassus. In 67, he was given another proconsulship under the lex Gabinia to clear the Mediterranean of pirates, and his imperium was extended further the following year by the lex Manilia, which authorized him to carry on the war in the East against Mithridates in Pontus and Tigranes in Armenia. He returned to Italy in 62 but continued to hold the proconsulship until his triumph in 61. In 57, he was given proconsular power to reorganize the grain supply of Rome and he continued to hold the post until at least 54, concurrently, in 55, with his consulship which he held with Crassus again as his colleague. Another law, the lex Trebonia, carried during their consulship, gave him a five-year proconsular command in the provinces of Nearer and Further Spain to suppress rebellions there, but he was allowed to govern the provinces through legates and remain in Rome where the Senate was preoccupied with rioting and the impending crisis with Julius Caesar. He continued to hold the proconsulship even in 52 when he was made sole consul, and he was still holding it in 49 when Caesar invaded Italy. He was then authorized by the Senate to share the command in the war against Caesar with the consuls, but at the end of the year he was given sole command of all the Republic's forces – the first time that a proconsul had been given supreme imperium in all theatres of war.³⁶

If the hypothesis that the M of our tile stamp stands for Mucia is correct, the P recorded on it should refer to Pompey's extended proconsulship in the 60s before he divorced her. It is of no great importance that Tile no. 1, the only one to be found in a stratified context, is dated to Phase I of the villa, which supposedly began around the middle of the first century BC, since the date cannot be pressed. The excavators tell us that the structures of Phase I are difficult to date because buildings of subsequent phases were built directly on top of the Phase I walls and many of the Phase I construction trenches were disturbed by later renovations. The date is based on the residual material, principally early Italian terra sigillata and late grey-gloss and black-gloss pottery, found in later phase fills, middens and construction trenches; and a small amount of the material recovered in the fill beneath the Phase I floors in two of the rooms suggests that the phase may date to as early as the second century BC (McCallum et al., 2011: 37).

Procuratrix/Procuravit?

An alternative explanation of the abbreviation P is that it relates not to Pompey but to Mucia, that it stands for procuratrix (or for the genitive procuratricis) and refers to the role which she held in the administration of her husband's properties during his long absences on campaign, according to the hypothesis put forward above. The use of the letter P as an abbreviation for procurator is well attested in inscriptions of the imperial period when the term was applied to high-ranking officials in the imperial administration, but the abbreviation is not otherwise known in the Republican period. Nevertheless the use of procurators by private individuals in the first century BC is well attested in literary texts. Our main authority on this is Cicero who frequently used trusted friends as procurators to carry out various types of business for him (Crook, 1967: 238). He gives a definition of the word in a speech composed, probably, in 69 BC in support of Aulus Caecina, who had been forcibly ejected from the property he claimed to have inherited by armed men acting on behalf of a rival claimant, Sextus Aebutius. Cicero asserts Aebutius’ culpability with a rhetorical flourish in the first person: ‘it makes no difference in law, at least in this kind of matter, whether a man who has forcibly ejected me is said to be your procurator, defined legally as someone who is effectively the owner of all the possessions of a man who is not in Italy or is absent on official state business, that is to say he is the deputy of someone else who is legally responsible, or whether he is your tenant farmer or neighbour or client or freedman or whoever, who has exercised force and ejected me at your request and in your name’.³⁷ The role of the procurator envisaged here, as the deputy of someone else who is legally responsible (vicarius alieni iuris) and who is absent from Italy, is exactly the role which Mucia must have had if she administered Pompey's estates in his absence. But could a woman be officially the procuratrix of someone else's possessions? The term is used by Cicero, but to refer to wisdom, sapientia, personified as the agent/protrectress of mankind (Fin. 4.7.17: cum sapientiam totius hominis … procuratricem esse vellent), not to a living woman. There is no other evidence for the use of the term procuratrix until the Late Empire.³⁸ But the verb procurare was occasionally used in non-juridical texts of women who organized matters in a domestic context.³⁹ That might also include the administration of Pompey's private affairs. It would also be possible in this sense to expand the P of our tile stamp to p(rocuravit) implying that Mucia administered the tile works on Pompey's behalf.

It cannot be claimed that either of these solutions, proconsul or procuratrix, is definitive, only that they are plausible interpretations. Without other evidence there can be no certainty as to the interpretation of the sigla P, which could have related to Pompey in several ways.

A tile stamp from the villa at San Gilio

These are not the only tile stamps that can be interpreted as referring to Pompey as a landholder in South Italy. Another from the Roman villa site at San Gilio in the upper Bradano valley, published by Helga di Giuseppe (2007: 171 and 170 fig. 8, no. 6; 2008: 351 and fig. 57), can also be claimed as Pompey's (Fig. 6). The letters, 1.1 cm high, are much smaller than those in the stamps from San Felice, but the letter forms are similar. There is probably a letter missing in the blurred right edge of the stamp which Di Giuseppe reads as Cn(eus) Ma(---). She regards the individual as unidentifiable, perhaps the owner of a figlina, but in the light of the other stamps discussed above it is probable that it too refers to Pompey. If so, it would belong to the first phase of occupation of the villa, which lasted from around the time of Sulla to the last decades of the first century.

Fig. 6. Tile stamp from the villa at San Gilio. Photo courtesy of Helga Di Giuseppe.

It is less certain that a tile stamp POMP impressed on a tegula fragment from the territory of Luceria refers to Pompey since the text could be expanded in several ways,⁴⁰ and the fact that it follows a different form of abbreviation from the pieces discussed above makes it unlikely that it stands for Pompeius Magnus.

The economics of Pompey's landholding at Botromagno/Vagnari

Neither the villa at San Felice nor that at San Gilio displays any of the lavish features that one would expect to see if they had been visited by Pompey himself. They can only have been inhabited by the staff who managed these estates. Their importance to him is likely to have been that they lie close to the traditional drove road which linked the winter pastures on the plain of Taranto with those in the Lucanian Apennines near San Gilio in the upper Bradano valley (Di Giuseppe, 1996; Small and Small, 2022: 162–3). The map Fig. 7 shows the line followed by the drove road shortly before it was abolished in the agricultural reform of the 1950s, but its course was dictated by the terrain and there can be little doubt that sheep were driven along much the same route in the Roman period. It is likely, therefore, that Pompey was investing in sheep-ranching along with his contemporaries Varro and Publius Aufidius Pontianus Amiterninus, mentioned above, and no doubt numerous other Roman senators and equites. But Pompey's sheep-ranching interests were on a grander scale, extending to the eastern Mediterranean where his son Gnaeus raised 800 men from his own slaves and herdsmen and brought them to Thessaly from Alexandria before the battle of Pharsalus (Caes., B Civ. 3.4). Back in Italy, Pompey himself armed slaves and shepherds and provided them with horses, creating a troop of 300 cavalry out of them before he crossed the Adriatic from Brindisi (Caes., B Civ. I.24.2). If, as seems likely, they were drawn largely or wholly from his own Apulian estates, their number gives us some idea of the size of his Italian stock-raising interests, which were probably of less importance to him than his eastern enterprises (Grelle et al., 2017: 28–30).

Fig. 7. Map showing the find-spots of stamped tiles discussed in the text in relation to the traditional drove road. 1 = San Gilio; 2 = Masseria Ciccotti; 3 = San Felice; 4 = Site 704; 5 = Botromagno (Silvium).

The evidence for Pompey's involvement in the exportation of wine from his Campanian estates has already been noted. The vast extent of his clientela in both East and West gave him unrivalled opportunities for exploiting his Italian estates for wealth creation.

Pompey and Silvium

The land where our three tile stamps were found, including the villa at San Felice and our Site no. 704, must have formed part, historically, of the territory of the Peucetian settlement known to the Romans as Silvium. It is likely to have been expropriated by the Roman state as public land, ager publicus, either at the end of the fourth century when they took the Peucetian city by storm, or in the Hannibalic War. The settlement declined drastically after that war, but it was refounded in the late second century as a village centred on a small villa, linked to Rome and Tarentum by the Via Appia (Small, 2020; Small and Small, 2022: 170–5). Since other smaller settlements were founded or refounded in the surrounding countryside, it seems likely, though it cannot be proved, that the agricultural territory needed to maintain the community was reconstructed out of ager publicus at this time.

But the revived settlement of Silvium was short-lived. Archaeological evidence shows that most of the village was abandoned, probably in the decade 80–70, and a large number of sling-shots found on the site suggests that it was captured in an assault, most probably in the war of Spartacus (Small, 2020; Schinco and Small, 2020). In the last phases of the war, Pompey, who had been campaigning against the Sertorians in Spain, was recalled with his army by the Senate and instructed to reinforce the consuls in the war against Spartacus and his followers. He arrived after Crassus had already won a major battle, but in time for mop-up operations. If these took him down the Via Appia (which is likely) then he will have passed the newly destroyed settlement of Silvium (Fig. 8). He perhaps seized the opportunity to buy up the deserted land at a bargain price, and then converted it into rough grazing for transhumant sheep.

Fig. 8. The Via Appia between Venusia and Tarentum showing places mentioned in the text.

The tile stamped CAESAR ꜸG, said to have been found on Botromagno, suggests that what was left of Silvium came to be owned by Augustus. How then did it pass from Pompey to Augustus? In late March of 44, Gaius Octavius, the future Augustus, who was studying at Apollonia in Epirus heard of his great-uncle's assassination and crossed the Adriatic to the coast of the Salentine peninsula. He then moved to Lupiae (Lecce) and waited there while he gathered support from Caesar's veterans before moving overland to Campania. In all probability he took the main route – the Via Appia – from Brindisi through Tarentum and Venusia, passing by the remains of Silvium and the countryside around (modern) Vagnari, where he would have seen the use to which Pompey had put the vast estate that he had created there. When he reached Rome in late April or early May 44, he is likely to have found that some of Pompey's confiscated estates were still up for auction. It is not too fanciful to suppose that he seized the opportunity to buy it, thereby making Silvium/Vagnari one of the earliest acquisitions for the imperial patrimonium. The fact that Cicero makes no mention of Octavian acquiring any of Pompey's estates is hardly significant. He had no wish to make an enemy of him, and, as we have already seen, his ‘lapse of memory’ in listing the purchasers in his Thirteenth Philippic excused him from doing so.

Others of his entourage are likely to have followed his example, including Vedius Pollio, one of whose tile stamps has been found in the remains of the villa at San Gilio, where the tile which can be attributed to Pompey was also found. They probably indicate that in the Augustan period the villa was owned by Vedius who must also have owned the neighbouring villa at the Masseria Ciccotti, where at least twenty tiles stamped P⋅V͡EI⋅POLLION have been found (Gualtieri, 2000: 332; 2003: 188; 2008: 216). What is known of his career has been pieced together by Syme (1961). He appears first as an anomalous equestrian governor of Asia, who is recorded in an inscription from Ephesus of the Claudian period (CIL III 7124) as the author of a constitutio which regulated the financial affairs and the administrative structure of the province, probably in 31/30 BC.⁴¹ To rise so far as an eques he must have already been an amicus of Octavian. He may also have been a business partner in the movement of transhumant sheep between winter pastures in the Basentello valley and summer pastures in the Lucanian mountains, following the example already set by Pompey.

There were, no doubt, many other supporters of the triumvirs who profited from the sale of Pompey's estates and created the socio-economic structure of Italy in the early principate.