Neighbours of the Apsaros Fort. Local Tribes on the Black Sea Coast during the Principate

Abstract

In the second half of the first century ce, the Romans built a fort at the mouth of the river Apsaros on the coast of Colchis. A Roman garrison was stationed there also in the second century and first half of the third. One of the reasons for fortifying the estuary of the river, given by both Pliny the Elder and Arrian, was the immediate vicinity of the kingdom of Iberia. Both Roman authors also described the local tribes living on the coast between Trebizond and Apsaros and further north. One wonders whether they were the indigenous population of the region and what kind of a relationship they had with the Roman Empire. This study searches for answers to these questions in the preserved written sources and in the archaeological record.

Introduction

The fort built by the Romans in the estuary of the Apsaros river (Çoruh River) (Fig. 1) has undergone archaeological excavations of varying intensity in the past years (Figs 2:1 & 3) and the work has also included, although to a much lesser extent, its nearest vicinity as well as the more outlying area on the banks of the Apsaros river in its lower reaches. Some of the exceptional finds, like the so-called Gonio hoard, have been published (Lordkipanidze et al. 1980; A. Kakhidze et al. 2015, 63–89). The Georgian–German expedition of 2000–2002 has also published the results of its investigations (Geyer 2003). The academic world is aware of other archaeological finds from the area, including some hoards and inhumation cemeteries from the Roman period, but there has never been a comprehensive study of the microregion bringing to the fore the issue of local settlement in this territory. Neither has there been exhaustive discussion of the archaeological findings in correlation with the knowledge of indigenous peoples passed down in their writings by ancient authors like Strabo, Pliny the Elder and Arrian. The issue has been treated very summarily in the margins of some publications (Chandrasekaran 2013; E. Kakhidze 2008, 313; Voronov 1974, 74–85; cf. Plontke-Lüning 2003, 7–13; 2005).

Figure 1. Map of the Black Sea in the Principate. Roman garrisons in Greek cities and forts on the coast of Scithia, Taurica and Colchis and important ports in the Roman provinces on the coast of Pontus: Byzantion, Sinope, Trebizond (Trapezus). (Drawing: O. Kubrak.)

Figure 2. Fragment of a 5-verstes map of the Caucasus region from the year 1883. Note the fort in the village of Goniya. Also marked: the southern arm of the Chorokhi delta (north of the fort) and the swampy areas north of the main river. (1) outline of the fort; (2) Makho village.

Figure 3. Apsaros (Gonio, Georgia) – Trajanic fort (Phase 2). An attempt to reconstruct the extent of fortifications and internal buildings: (1) porta praetoriana; (2) headquarters (principia); (3) porta principalis sinistra – probable location; (4) garrison bathhouse (balneum); (5) course of moat; (6) extent of Late Roman and Byzantine walls. (Drawing: O. Kubrak.)

Analysing the scattered information on local tribes inhabiting the area at the mouth of the Apsaros river in the first centuries ce, one is led to the understanding that the Roman fort at Apsaros was constructed on ground that belonged to two tribes in particular, the Heniochi and the Macrones. Despite differences and doubts, linguists usually categorize both peoples among the group of western Georgian tribes, indicating a very probable linguistic affinity with modern-day Abkhazian, Svan or Megrelian. A synthetic picture of these peoples and the local society they formed emerges from a survey of the ancient sources in which they were mentioned, understandably with inevitable Roman bias. These data, critically reviewed and put next to an overview of the diverse archaeological findings from the region, give a better understanding of the social and cultural environment in which the Roman fort of Apsaros functioned from the middle of the first century ce through the mid-third.

The indigenous people of the Colchis coast in ancient written sources

The indigenous tribal landscape of the region on the Colchis coast in the Roman period is presented in a number of sources, the most important of which are the works of Pliny the Elder and Lucius Flavius Arrian. These texts add information to the lists of peoples and places in the Caucasus and eastern frontiers of Asia Minor known from Greek and Latin texts written long before the appearance of the Romans on the Black Sea coast (collected and published already in the nineteenth century and the early twentieth; see Gan 1884; 1890; Latyshev 1890; 1904; 1906; 1947; 1948a,b).

The two Roman authors, whose texts are of greatest importance for getting to know the barbarian tribes living around the Roman forts on the Colchis coast, augmented information taken from some older sources with new knowledge current in their times and, in Arrian's case, personal observation. Moreover, these two accounts are well dated. Pliny listed the forts of Apsaros and Sebastopolis, which could not have been constructed before the annexation of Polemon's kingdom of Pontus in 64 ce, while we can be sure that the information in his text predates his death in 79 ce (Fig. 4). As for Arrian's inspection trip as governor of the province of Cappadocia, it took place in 131 or 132 ce (see Speidel 2009, 603–4; Wheeler 2011: 126). A comparative analysis of the two sources leads to some very interesting conclusions concerning the dating of individual Roman garrisons (Fig. 5) (see Karasiewicz-Szczypiorski et al. 2019). While there are some differences between the two sources regarding the names of tribes, rivers and settlements that are listed, they are quite coincident on the whole.

Figure 4. Coast of Colchis in Nero's time. Territories of the Heniochi and Macrones tribes. Greek colonies: Trebizond. Roman forts: Apsaros and Sebastopolis. (Drawing: O. Kubrak.)

Figure 5. Coast of Colchis in the times of Trajan and Hadrian. Territories of the Heniochi and Macrones tribes. Greek colonies: Trebizond. Roman forts: Hyssou Limen, Apsaros, Phasis and Sebastopolis. Abandoned forts: Athenai. (Drawing: O. Kubrak.)

The two descriptions referring to the coast of Colchis in the second half of the first century and first half of the second have been broadly studied (including commentaries to critical editions; see the list of editions of primary sources used by the author), but without undertaking a detailed analysis of tribal borders for the different peoples said to have inhabited the coastal regions east of Trebizond and west of the Apsaros river. It could hardly have been a coincidence that one of the first Roman forts in the region was constructed at the mouth of this river. The results of recent geoarchaeological coring at Gonio have confirmed the presence of a harbour, which made use of a body of inland water connected to the sea. It was located on the southern fringes of a coastal zone where the city of Batumi is today, in antiquity forming an extensive river mouth (delta) enclosed by hills coming down to the sea (Łęczyński et al. 2019). The mountains appear to have been respected as a distinct natural boundary by barbarians and Romans alike.

East of Trebizond, according to Pliny, there was the tribe of the Sanni. He described them as the ‘Heniochi people’ (gens Sannorum Heniochorum). Beyond them lay the river Absarron and a fort of the same name, and beyond the mountains was Hiberia (lands under the control of the king of Iberia). Other people lived on the coast, but the first in line to be mentioned by Pliny are the Heniochi (Plin., HN 6.4. 12). Since Pliny then speaks of another river, the Phasis, which he considers as an important boundary, this suggests that the Heniochi could have been the closest neighbours looking from Apsaros (Fig. 4). Arrian's much more detailed description of the stretch of coast in question gives more names of tribes, settlements and smaller rivers flowing into the sea. The Sanni are also named as Trebizond's neighbours to the east, but Arrian identifies them explicitly with the Colchians. To his knowledge, the territories of the Macrones and Heniochi, tribes ruled by a king named Anchialus, lie east of the tribal land of the Sanni (Arr., Peripl. 11) (Fig. 5). This ruler resided about 7 km (40 stadia) east of the fort of Athenai, at a place located at the mouth of the river Prytanis; the fort itself was already deserted in Arrian's time (Arr., Peripl. 7). The Zydretae, who were the subjects of Pharasmanes, king of Iberia, were the neighbours of these two peoples (Arr., Peripl. 11).

The kingdom of Iberia lay beyond the line of hills to the east of the Apsaros fort (Braund 1994, 185). The border between the two states could have continued in the hills north of the Apsaros delta, placing the territory of the Zydretae further north and on the coast (cf. Barrington Atlas, 87, Pontus-Phasis; Bosworth 1993, 250; Juntunen 2013, 156; cf. also: Anon., Peripl.42).

The relations of the Kingdom of Iberia with the Roman Empire varied in intensity and character, especially during the reign of successive Antonine emperors. King Pharasmanes II, a contemporary of Hadrian and Antoninus Pius, was a very difficult partner, manifesting great friendship for the current emperor when needed and quickly striking new alliances if the tide of regional politics changed (Braund 1994, 232–3). Unsurprisingly, Trajan pragmatically decided to reconstruct the fort at the mouth of the Apsaros, a decision upheld by Hadrian who subsequently expanded the fortifications.

The Sanni tribal territory closest to Trebizond was at this time strongly connected to the province of Cappadocia. Arrian writes that the natives had no king and paid (or at least had previously paid) tribute (φόρος). A fort manned by an infantry cohort, overlooking a commercial harbour, stood at Hyssou Limen (Arr., Peripl. 3) (Fig. 5). Arrian noted that the natives were battle-hardened and hostile towards neighbouring Trebizond and also inclined to banditry (Arr., Peripl. 11).

In view of this information, it can be assumed that Anchialus’ realm encompassed the land lying between the sea and the Pontic mountains, extending from the country of the Sanni in the west to the Apsaros river in the east (Fig. 5). The late antique anonymous Periplus (Anon., Peripl. 49 and 51), partly based on Arrian's text, confirms this assumption. In turn, the fort of Apsaros controlled not only a convenient sea and land route leading from the north to the province of Cappadocia, but also separated the territories of Rome's allies, the Heniochi and Macrones tribes, from those of the Zydretae who were subordinated to the kings of Iberia. Based on these assumptions, any evidence of local settlement from the first centuries of the common era found in the lower reaches of the Apsaros river can be associated with the Heniochi (Pliny) or the Heniochi and Macrones (Arrian).

The Heniochi

Pliny mentions gens Sannorum Heniochorum and Heniochi in the area between Trebizond and the Apsaros river (Plin., HN 6.4.12), but he also refers to the Heniochi in connection with places located much further north, i.e. in the vicinity of the towns of Dioscurias and Pityus (Plin., HN 6.5.16). The former was said to be founded by the charioteers Amphilus and Telcheius, participants in the Argonaut expedition (Fig. 6). The legend had it that the indigenous people living in the area were named Heníochoi from the Greek (Ἡνίοχοι, ‘charioteers’). Pliny may have taken this legend from an unpreserved work of Artemidorus, later cited also by Strabo (see RE VIII/1: 260, 20; cf. also Kolendo & Płóciennik 2015, 124) and repeated in several other sources, including Pomponius Mela (Pompon. 1.111), Charax of Pergamum (Eust. comm. ad Dionys. Per. 687), C. Julius Solinus (Solin. 15.17), Ammianus Marcellinus (Amm. Marc. 22.8.24) and the medieval writer Eustathius (Eust. comm. ad Dionys. Per. 680–687), who wrote commentaries to the works of Dionysius Periegetes. In another part of his work, Pliny refers to these peoples as the Heniochorum plurima genera, mox Achaeorum (Plin., HN 6.12.30). The plural form, suggesting a group of tribes, is important in this passage (Fig. 6), as is also the hint that they were the neighbours of the Achaeans on the Pontus.

Figure 6. The coast of Colchis before the fall of the kingdom of Mithridates VI Eupator. Territories of the Heniochi and Macrones tribes. Greek colonies: Trebizond, Phasis and Dioscurias. Other ports: Pityus. (Drawing: O. Kubrak.)

It is intriguing that Arrian fails to mention the northern faction of the Heniochi. Perhaps the governor of Cappadocia listed only tribes allied with Rome, and especially those whose rulers held their kingdoms by imperial conferral?

In other, earlier texts, the Heniochi are listed among the tribes inhabiting the Pontic coast along the western slopes of the Caucasus, either going south from the Bosporus and Sindica or northward (Figs 1 & 6; Table 1). Sources detailing Mithridates’ escape following his defeat during the war against Pompeius at Bosporus have the king moving overland along the Pontus coast (Braund 1994, 158; Coşkun 2021, 250), passing successively through the lands of different tribes depending on the source: the Colchians and the Heniochi (Livy, Per. 101), the Heniochi and the Zigs (Str. 11.2.13), the Heniochi and the Achaeans (App., Mith. 102). Chasing the fugitive, Pompeius was said to have fought against the Colchians, the Heniochi and the Achaeans (Vell. Pat. 2.40.1). The Heniochi and the Achaeans were also listed among the peoples whom the Roman commander conquered and which he presented during his triumph in Rome (App., Mith.116).

Table 1. A compilation of the names of indigenous tribes living on the Colchis coast mentioned by ancient authors.

Another fleeing king, Vonones of the Parthians, was said by Tacitus to have escaped from Pompeiopolis in Cilicia, where he had been detained by the Romans, to his cousin the king of the Scythians, via the lands of the Armenians, the Albanians and the Heniochi (Tac., Ann. 2.68). The Heniochi in this case could be assumed to be a tribe living in the western Caucasus. Finally, Stephanus of Byzantium wrote that Inioheia was the name of a country in the Caucasus (Steph. Byz. 302).

It appears from the above that while the ancient authors were in agreement that the tribe (or group of tribes?) referred to as the Heniochi inhabited the area by the Black Sea in the western Caucasus long before the arrival of the Romans, none of the sources ever suggested that they were living east of Trebizond before the time of construction of the first Roman forts (Fig. 6).

The Macrones/Machelones

Some of the sources mentioning the Heniochi also referred to another tribe, the Machelones/Macrones, whose lands lay further to the south. The earliest information about these people can be found in a text by Hecataeus (Hecat. 191), who associated them with the Sanni, Trebizond's eastern neighbours. Strabo repeated this information (Str. 12.3.18), as did also Stephanus of Byzantium (Steph. Byz. 439). However, only Arrian identified them with the Macrones (Arr., Peripl. 11).

In his considerations of circumcision, which was supposed to have originated with the Egyptians, Herodotus (Hdt. 2.103) says that it was also practised by the Colchians, who had adopted the custom from the Macrones. These tribes were obviously living next to one another. Herodotus also lists the Macrones among the peoples who paid tribute (Hdt. 3.94) to the King of Kings and sent a contingent to serve in his army (Hdt. 7.78). Xenophon's Anabasis (Xen., An.4. 8. 1–3 and 4.8.16–24) is a very important source—because based on personal experience—confirming the placement of the Macrones east of Trebizond (Fig. 6). In the times of this Greek historian, the Macrones and the Colchians were supposedly free peoples, not subordinated to the Persian king (Xen., An. 7.8.25).¹ Diodorus Siculus, who also mentioned the Macrones, took his information from Xenophon (Diod. Sic. 14.29.5).

The Periplus (Ps.-Scyl. 85), the original version of which was supposedly edited by Scylax of Caryanda, does not mention the Macrones in the vicinity of Trebizond, but it does list the Macrocephali (Makrokefaloi’) (Manoledakis 2022, 209–10). Later authors also linked these two names together (schol. ad Ap. Rhod. 2.1242). Hippocrates wrote about the rite of skull deformation practised by this mysterious tribe (Hippoc., Aer. 14).

The Macrones are mentioned in some texts listing the peoples inhabiting the coast in the vicinity of Trebizond and further east in the direction of the Phasis river, but not as neighbours of the Heniochi (Amm. Marc. 22.8.21; Dionys. Per. 761–774; see also Eust., comm. ad Dionys. Per. 762 and 765 and Priscian., Per. 740).

Pliny's mention of the ‘Macerones’ inhabiting the area up to the Absarron River (Plin., HN 6.11.29) appears next to information about the numerous tribes of the Heniochi in the western Caucasus and on the banks of the Pontus (Plin., HN 6.12.30). In another place, he writes of the Heniochi in the vicinity of Apsaros (but not at the sametime as the Macrones) (Plin., HN 6.4.12) (Fig. 4).

Considering the different information contained in descriptions of the Caucasus, one is led to think that it must have originated from different sources and might have referred to different times.

In Arrian's account, the Macrones (Machelones) are said to have occupied, together with the Heniochi, the coast east of Trebizond (Fig. 5) and west of the Apsaros river (Arr., Peripl. 7 and 11), information later included in the late antique anonymous Periplus based on his account (Anon., Peripl. 49 and 51). The latter source simultaneously lists the Heniochi further to the north, describing them as a tribe inhabiting land between the western Caucasus and the eastern coast of Pontus (Anon., Peripl. 59).

The Heniochi west of the Apsaros river

Summing up this comparison of source texts, we can assume with a high degree of probability that the Heniochi, actually a group of tribes, were originally settled in the southwestern part of the Great Caucasus, neighbouring the Achaeans living further to the north on the coast (Fig. 6) (Plin., HN 6.12.30). The Macrones had their ancestral seat east of Trebizond. Pliny was probably the first to place both peoples in the same area, while Arrian added the information that the two tribes (or tribal groups) were united under one ruler (Arr., Peripl. 11). The anonymous Periplus, which is largely based on Arrian's account but is a compilation of various sources, contains an intriguing passage that the Machelones and the Heniochi inhabited territory from the Archabis river to Ophiuntus, once occupied by the Ekhiri tribe (Anon., Peripl. 42). While it is impossible to link this information to a specific time, it does show an awareness of population changes having taken place in the area in question. Were the Machelones and the Heniochi a completely new union of tribes (perhaps with the participation of an immigrant population)? Or was this simply evidence of expansion at the territorial expense of their neighbours?

The theory about the territorial expansion of the Heniochi that was developed based on this piece of information (RE VIII/1, 273, 44) suggests a migration taking place from the north (from the Caucasus) into territory previously occupied by the Sanni and the Machelones. The date of such a migration is a crucial issue. Considering that Pliny mentions the Heniochi in relation to information about the fort on the Apsaros river, it can be assumed that these peoples already occupied new settlements in the area during Nero's reign (see Wheeler 2011, 133), although most probably they were not related to the Macrones living further west (Fig. 4). Pliny is also the source of the information that the Heniochi occupied lands previously belonging to the Macrones at the mouth of the Apsaros river. However, there is no sound base for a reconstruction of how this relocation took place.

Changing urban landscape

Pliny's description of the Colchis coast includes information about places that had been port towns and regional centres of exchange and trade in the period before Roman domination but which—Pliny makes it abundantly clear—no longer existed (Fig. 6). These were notably the Greek colonies, one of which was Phasis (at the mouth of the river of the same name), founded by the Milesians on land originally belonging to the Heniochi (Heraclid. Pont. 18). Strabo writes of Phasis as a town by the river of the same name, a Colchian market port situated between the river and the lake (Str. 11.2.17). Next along the coast was Dioscurias where, according to Strabo, numerous mountain tribes, over 70, (including the Soanni/Svans), came from the Caucasus to buy salt (Str. 11.5.6). Pliny wrote of representatives of 300(!) tribes speaking various languages visiting this city of Colchis (urbe Colchorum) to conduct business there with the help of 130 translators (Plin., HN 6.5.15). Notably, Pliny used the term ‘urbs’ and not ‘oppidum’ only in reference to this centre. Its prosperity and dominant position is also borne out indirectly by the fact that King Mithridates VI Eupator, fleeing from Pompey's army in Bosporus, chose to spend the winter of 66/65 bce in this town (App., Mith. 101; Coşkun 2021, 250). Pomponius Mela, writing later in 43–44 ce, also placed Dioscurias in the land of the Heniochi; he may have been referring to an earlier reality (Pompon. 1.111). Strabo also listed the great Pityus alongside Dioscurias (Str. 11.2.14), supposedly neighbouring the Heniochi coast (Fig. 6). Pityus may have been a trade centre with local roots. However, it cannot be ruled out that the first emporium was established by the Greeks. Another presumed harbour south of Phasis, on the coast of Colchis, was Bathys Limen/Portus Altus (Braund 1994, maps 2 and 3; Barrington Atlas, 87, Pontus-Phasis).

Pliny describes Dioscurias as a deserted city (Plin., HN 6.5.15) and Pityus, ‘the richest city’, as destroyed by the Heniochi (Plin., HN 6.5.16). He points out that many cities had once been located on the banks of the Phasis river, including the city of Phasis at its mouth (Plin., HN 6.4.13). Writing about rivers flowing into the Black Sea, he includes the Phasis and Bathys, but is silent about any cities/ports on their estuaries (Plin., HN 6.4.4). Instead, he mentions two Roman forts (castella), one on the Absarron River and the other called Sebastopolis (Plin., HN 6.4.12 and 14) (Fig. 4). He also refers to several other towns (oppida) on the coast. Counting from the south (from the Absarron River), these were: Mation (south of Phasis), Cygnus and Penios (both in the same area as Sebastopolis), then Heracleion (between Sebastopolis and Pityus) and Hiero (in the land of the Achaeans). Travelling along the same shore about half a century later, Arrian saw none of these places. It is possible that they were makeshift marketplaces/settlements with harbours, whose short-lived existence was linked to the sudden collapse of the old urban centres.

That these traditional centres of trade were not reconstructed in the early second century is clear from Arrian's account. Writing about Sebastopolis, Arrian added that the place had formerly been called Dioscurias and that it was once a colony of Miletus (Arr., Peripl. 10), while Dioscurias itself was for him only a fort (Arr., Peripl. 17). At the mouth of the Phasis river, he mentions temples of a local cult and a fort and adjacent harbour with a settlement, with army veterans among the inhabitants (Arr., Peripl. 9). No mention is made of an earlier town. Pityus, which lay beyond the reach of Roman rule, was the first convenient harbour on the way to Bosporus (Arr., Peripl. 18). It also seems that the new oppida mentioned by Pliny did not survive (let alone gain in importance). Based on the description of Phasis, it can be assumed that the role of ‘ports of trade’ was gradually taken over by camp settlements developing around the forts. Similar settlements probably existed at that time outside the walls of Apsaros and Sebastopolis, although Arrian does not mention them. The later establishment of another fort at Pityus, which had additional external fortifications (like Phasis), may also have been intended to create a safe space for merchants.

The destruction of the all-important port cities on the Colchis coast and in the navigable lower reaches of the Phasis river coincides with an apparent migration of a part of the Heniochi people from the Pontic coast at the foot of the western Caucasus to the coastal zone further to the south, that is, on the Apsaros river (Fig. 4). These events cannot be dated precisely, but they occurred sometime after the fall of Mithridates (63 bce) and before the annexation of Polemon's Pontus (64 ce). It would have been a gradual process, related to the destabilization of the region after the disappearance of the hegemonic kingdom of Mithridates.² At least two large armies crossing the eastern coast of Pontus during this period could have been responsible for this massive crisis: the Bosporan army of King Pharnaces II marching south (48 bce) and the army of King Polemon I of Pontus heading toward Bosporus (14 bce). However, the information that Pityus was destroyed by the Heniochi (Plin., HN 6.5.16) may indicate some involvement of these local people also in the devastation of other ports and harbours.

Sailors and pirates

The Heniochi appear also in accounts of their cruelty toward their enemies, as well as banditry and sea piracy. Behaviour of this sort was supposed to be common among the tribes living at the foot of the Caucasus, on the eastern Pontus coast; it was also described as typical of the Crimean Tauri. Several authors mention the cruel and savage inhabitants of the region: Diogenes of Sinope writes about the Heniochi and the Achaeans (Diog. Sinop., Epistolai), so does Aristotle (Arist., Pol. 8.4.4), while Dionysius of Halicarnassus mentions the Achaeans, the Zigs and the Heniochi (Dion. Hal., Ant. Rom. 1.89). Sallust refers to the Achaeans and the Tauri (Sall., Hist. 3.48), as do the scholia to Junius Juvenalis (schol. ad Juv. 15.115), whereas Eustathius’ commentary to Dionysius Periegetes (Eust., comm. ad Dionys. Per. 687) lists the Achaeans, the Heniochi and the Zigs. The Colchians and the Heniochi appear among the peoples personifying evil in the Sibylline Oracles (Or. Syb. 12.55). In this case, it is probably a trope, the source of which might be a fragment of The Jewish War by Titus Flavius Josephus. The ancestors of the Heniochi were also said to have been cannibals and to have skinned people (Arist., Pol. 8.4.4; Heraclid. Pont. 18).

None of these numerous references specify what these savage tribes did for a living, but it is clear from other accounts that they must have engaged in sea banditry. Diodorus Siculus mentions the Heniochi, Tauri and Achaeans practising piracy in the Black Sea (Diod. Sic. 20.25.1).³ Strabo (Str. 11.2.12) records that the coastal dwellers—the Achaeans, the Zigs and the Heniochi—engaged in sea banditry, for which they used boats called kamarai. They ruled over the sea, captured slaves, raided areas under Roman administration and sold their loot (probably also slaves) in Bosporus. His account is crucial to understanding the situation in the eastern Pontus at the turn of the era (cf. Wheeler 2011, 133–4). Ovid, who lived in Tomis at that time, in one of his ‘Letters from Pontus’ (Ov., Pont. 4.10.25), recorded that the Heniochi and Achaean sailors were more terrible than Scylla and Charybdis. News of rampaging pirates far to the east reached the poet even though he lived in a city threatened daily by the incursion of nomads from the Scythian steppes.

The inclination to sea (and maybe also land?) banditry was not a new thing. According to Appian, Pompey was said to have dealt with bands of robbers even as he fought the war with Mithridates (App., Mith. 114) and, more importantly, his victories over the Heniochi and Achaeans were later gloated over during his triumphs held in Rome (App., Mith. 116). It cannot be excluded, however, that the battles against the Colchians, the Heniochi and the Achaeans were the result of at least some of these tribes coming out in support of the fleeing Mithridates (Coşkun 2021, 250) (Str. 11.2.13; Vell. Pat. 2.40.1; Luc., Bell. civ. 2.590).

The annexation of the Pontic kingdom of Polemon II and the reorganization of his royal fleet into a Roman provincial one (in 64 ce) brought some peace and order to the region, but not for long. The region was destabilized again by the Anicetus revolt, which broke out in 68–69 ce. The first Roman forts on the Colchis coast—that is, Apsaros and Sebastopolis (cf. Wheeler 2011: 140)—were built after 64 ce and stability returned in the 70s (Fig. 4). It seems that the location of these forts was determined by multiple factors, namely, the proximity of convenient harbours and the potentially hostile presence of the kingdom of Iberia, which remained outside the Empire. However, perhaps both forts were located in such a manner primarily to retain control over areas inhabited by the Heniochi? The Roman military presence in Colchis would thus have stemmed primarily from the need to combat maritime banditry.

A passage from a speech by Herod Agrippa II, said to have been delivered to the inhabitants of Jerusalem in 66 ce, shortly before the outbreak of the First Jewish Revolt (cf. Wheeler 2011, 131–2), is probably an account of a recent widespread campaign against the pirates in Pontus. The Heniochi and the Bosporans are mentioned here alongside the more widely recognized Colchians and Tauri, showing the author of the speech to be well aware of local conditions. Since the speech was written for the king most likely by Titus Flavius Josephus, a participant or witness to many of these events and a person who was privy to the content of military reports and the private correspondence of King Agrippa and the Roman emperors Vespasian and Titus, its credibility with regard to historical detail is rather not questioned (Radożycki 2001, 31, 32, 33). It can be assumed that the Romans acted against the pirates immediately after the incorporation of the kingdom of Polemon II into the Roman Empire but before the outbreak of the Jewish revolt, that is, in the two years between 64 and 66 ce (contra Wheeler 2011, 133–4, 140). This action at a later time is unlikely because Roman forces in the east were engaged in fighting on the Jordan river, while in Pontus Anicetus had rebelled (68–69 ce). Tacitus reports that the revolt was joined by local barbarians using boats built by them for sea operations, while Anicetus himself, facing defeat, had fled to the safety offered him by one of the Colchis tribes (Tac., Hist. 3. 47–48). Thus, it must also be taken into account that the circumstances described by Flavius Josephus referred to the ‘Year of the Four Emperors’ when the revolt in Pontus was suppressed. It also appears that the Roman crackdown on pirates targeted Bosporus for its profiting from banditry. However, with epigraphic evidence scarce from the region (and dated to later times; see Saprykin & Ermolin 2010; Ivanchik 2013), it is difficult to point to sources that could confirm Roman intervention in this client state at this particular time. It is also probable, based on Josephus’ information, that piracy in Taurica was eliminated as part of the same campaign (see Wheeler 2011, 135). However, the issue of the presence of the Roman fleet and possible garrisons on the southern shores of the Crimea in the first century ce has long been debated. The arguments raised by the sceptics in this discussion still seem very strong (see: Sarnowski 2006a,b,c).

The burning issue of dealing with rampant piracy and ensuring the safety of maritime shipping lanes has found ample confirmation in the results of recent excavations at the Apsaros fort. The first fortifications and buildings inside the garrison were constructed in all probability during Nero's reign, coinciding presumably with the annexation of Polemon's Pontus (Karasiewicz-Szczypiorski 2016; 2018; Karasiewicz-Szczypiorski & Mamuladze 2019). However, the fort seems never to have reached a stage of completion as planned and was evacuated after a relatively short time, possibly during Anicetus’ uprising (Karasiewicz-Szczypiorski 2016, 62).

There is one other reference in the written sources worth noting in this context. Tacitus recounts the story of a Roman army returning westward along the coast of Taurica following an intervention in Bosporus: ‘During their withdrawal, however, fortune changed, as a few of the ships—they were returning by sea—were carried on to the Taurian coast and there surrounded by the barbarians, who killed the prefect of one cohort and many of the auxiliaries’ (Tac., Ann. 12.17–21). The expedition to Bosporus and the events that followed are dated to 45–49 ce (Zubar’ 1998, 33; cf. M. Novichenkova 2016, 221). A Tauric cult site discovered at the Gurzufskoe Sedlo pass in the Crimean Mountains (N. Novichenkova 1994; 2002), which has yielded offerings that include elements of weaponry, armour and other objects (Novichenkova 1998), dating to between the second half of the first century bce and the first half of the first century ce (N. Novichenkova 2014, 147; M. Novichenkova & Kontny 2015, 118), may reflect this staggering slaughter of Roman soldiers.

The losses suffered on this occasion, somewhere on the beaches of Taurica, may have triggered the retribution against pirates in Pontus. However, it has been difficult so far to provide evidence of Roman garrisons stationed in the Crimea at the same time as the forts at Apsaros and Sebastopolis were being built. The fort at Cape Ay-Todor near Yalta, which controlled a large part of the mountainous coast inhabited by barbarians and was crucial to the safety of water transport (the garrison most likely included a lighthouse built and operated by the army: Karasiewicz-Szczypiorski 2015, 93; 2019, 162), was almost certainly not built until the early second century ce. Thus, it corresponds in time to the reconstruction and expansion of the fort at Apsaros during the reigns of Trajan and Hadrian. Nothing in the archaeological record so far has confirmed the presumed simultaneous expeditions against pirates in northern and eastern Pontus in the 60s ce.

Indigenous people of the Colchis coast in the eyes of outsiders

The information we have about the everyday life, as well as the material and spiritual culture of the barbarian tribes in question, already extremely limited, is further filtered through the eyes of ancient authors who often succumbed to the urge to ply their readers with interesting curiosities. In any case, their information was mostly rehashed from earlier descriptions.

The Heniochi

Dionysius of Halicarnassus (Dion. Hal. 1.89) and Strabo (Str. 17.3.24), discussing the Achaeans, and the Zig and Heniochi tribes living on the eastern coast of Pontus, noted that these peoples led a nomadic life and one of banditry. One wonders to what extent this opinion was driven by the explanation of the origin of the name of the Heniochi tribe, which the Greeks derived from Ἡνίοχοι, Heníochoi, meaning charioteers, possibly because of an onomatopoeic play on how it sounded to their ears (Marr 1913, 325–31). It could also be a reflection on the nomadic-style carts (similar to that of the Scythians: Hdt. 4.46.3), which the Heniochi may have used to get around. This hypothesis, however, is difficult to prove and not very probable.

Considering that, at least according to Pliny (HN 6.12.30), the numerous Heniochi tribes lived in territory between the Svani territories on one side and the coast of Pontus on the other, then those who lived at some distance from the sea undoubtedly occupied various mountain habitats that would have determined a specific form of economy, pastoralism (including transhumance pastoralism), for example, which outsiders would have regarded as typical nomadism (Figs 4–6). This is all the more so as summer outings to the mountain meadows probably involved whole families accompanying their herds. Even today, when summer comes, livestock is driven to the mountainous Tusheti region in eastern Georgia. For the winter, the herds are brought down again to the lower-lying area of Kakheti. In light of contemporary observations (based on the author's own experience), the peculiarity of the Tusheti region (and of the Khevsureti as well) lies in the fact that these mountainous summer residences constitute the local communities’ homeland. It is difficult to say today whether it was the same in the western Caucasus in ancient times.

The information about numerous tribes coming down from the mountains to Dioscurias to buy salt (Str. 11.5.6) can be linked to a pastoral lifestyle. Salt is a product needed in large quantities for the animals of these pastoral communities, which in turn can sell their mountain produce or selected cattle in the city. The mechanism of a city becoming rich on the salt trade conducted with mountain shepherds has been described on the example of early Rome by Adam Ziółkowski (2004, 52–3).

The coastal brethren of the mountain Heniochi would have engaged in active sea banditry with the use of seagoing vessels. Strabo recalls boats called kamarai (Str. 11.2.12), which were supposedly narrow and light, easily taken ashore and concealed if necessary. Each of them could hold 25 men, and on rare occasions up to 30 (Eust., comm. ad Dionys. Per. 700). Therefore, the barbarian warriors sailed the sea in relatively small vessels. They would almost certainly have set out on expeditions in a group of such ships, also effective for river navigation due to their size. The most detailed description is offered by Tacitus, who writes of the indigenous people supposedly supporting the rebel Anicetus in Pontus in 68–69 ce. They sailed in boats they had built, which had two rudders and two bows, thus enabling motion back and forth without having to turn the vessel. The hull in this case was said to be wide and joined without the use of ‘bronze or iron’. The practice in bad weather was to form a superstructure on the port side with additional planks. That these are vessels of the same kind as reported earlier is proven by Tacitus’ use of the Latin term camarae for these boats (Tac., Hist. 3.47). Unfortunately, neither Strabo nor Tacitus recorded whether these were only rowing boats or perhaps used for both rowing and sailing. The chaikas that the Cossacks operated on the same waters in the seventeenth century (Beauplan 1660, 55–7, fig. H; Çelebi 2008, 179–87) may be a distant parallel, yet substantiating the information found in ancient accounts. The expediency and effectiveness of using small units that could operate in groups during looting expeditions on the eastern Pontus coast is hardly arguable.⁴

The Macrones/Machelones

The source accounts about the Macrones/Machelones are subject to the same reservations about their reliability as in the case of the Heniochi, with one exception, namely Xenophon, who wrote about them from personal experience. His account, written more than 450 years before this part of Pontus was incorporated into the Roman Empire (Xen., An. 4.8.3), spoke of woven (wicker) shields, spears and ‘hairy’ chitons. From this description we can infer that the shields resembled Greek pelts and the chitons were probably made of hairy cloth (rather than leather worn with the bristle side up). We can find more information about the shields in Xenophon's description of the neighbouring tribes with whom the Greeks were then at war. Along with some natives, about 20 shields were captured. They were also woven and covered with raw cowhides (with bristles) (Xen., An. 4.7). Herodotus also mentioned wooden helmets and small shields (although it is not clear whether wooden or woven and whether covered with skins), as well as small spearheads and long spears (Hdt. 8.78). Spears were also referred to by Diodorus Siculus in a fragment modelled on Xenophon (Diod. Sic. 14.29.5). The Greek historian reported that, upon concluding an alliance or peace, the Macrones conducted a symbolic exchange of spears with the enemy. It can be assumed, based on these mentions, that a man/warrior carried a shield and spears, either one or two (long and short). If the latter, then one would have been used for throwing and the other for hand-to-hand combat. The use of such weaponry and the combat method are confirmed for other communities in the European Barbaricum during the Roman period (Kontny 2001, 92, 102, 111–12; 2019, 42, fig. 23). Two warriors similarly equipped, one with a short and the other with a long spear, are depicted on a silver rython from the site of Gomi in western Georgia, dated to the first century bce–second century ce (Kruk 2024, 310).

Strabo speaks of the ‘Macropogones’ [Longbeards], interchangeably with the Moschi, who wore bushy beards and were neighbours of the coastal Heniochi living further to the south (Str. 11.2.1, 14). They could be somehow related to the Macrones. Herodotus writes that the contingent organized by the Macrones and Mossinoeci for the King of Kings had the same weapons as the Moschi (Hdt. 7.78). Thus, perhaps all these peoples, including the Heniochi, were not only similarly armed but were also similar in appearance, perhaps wearing distinctive beards. That they all had beards is, of course, pure speculation.

The Macrones have also been confused with the Macrocephali (schol. ad Ap. Rhod. 2.1242), a term signifying people with long heads. Hippocrates (Hippoc., Aer. 14) considered it a form of deliberate skull deformation. Hence, it can be assumed to be a custom adopted also by the Macrones living east of Trebizond, although it is more than likely that Xenophon would have noted it had this been true. Soil conditions at archaeological sites are not conducive to the preservation of bones and organic artifacts (the results of excavations at the inhumation cemetery at Pichvnari near Kobuleti is a good example: Kakhidze 2007, 211), hence this information has little chance of being verified in excavations. In this part of Pontus, local soil conditions have led to the decomposition of bones and organic objects at archaeological sites.

Another intriguing piece of information, unconfirmed by other written sources, comes from Herodotus, who says that the Macrones adopted the custom of circumcision from the Colchians (Hdt. 2.103). This change, according to Herodotus, occurred ‘not long ago’. It can be assumed to be a somewhat unjustified inference resulting from a legend about the Colchians being descended from the Egyptians. Circumcision practised in both communities was supposed to constitute proof of such origins. The Macrones (in the pre-Roman period) are sometimes described as the neighbours of the Colchians (Xen., An. 4.8.6-9 & 7.8.25; Diod. Sic. 14.29.5); thus, attributing such a ritual to them could be the result of an assumption that peoples living next to each other tend to have similar customs.

The Heniochi and the Macrones in the archaeological record

More could be said about local settlement of the Roman period in the territory between modern-day Trabzon and Batumi had investigations been conducted on a larger scale in this region. So far, however, discoveries have been limited to a small area within the borders of Georgia, mainly on the banks of the lower Çoruh River, a short distance from the Apsaros fort (Figs 4–5). One group of finds consists of inhumation graves and the grave goods found with the skeletons. The other is a set of tantalizing hoards, the context of which is rarely known and the composition of which is often less than precisely published.

The two cemeteries, in Makho (Fig. 2: 2) (A. Kakhidze & Shalikadze 2010; 2015) and in Kapandibi (Ebralidze et al. 2010), present very similar grave assemblages which are dated no earlier than the third century ce by a collection of coins from the reigns of Aurelian and Diocletian (c. 270–294 ce) found in the graves at Makho (E. Kakhidze & Mamuladze 2014: 178; cf. Ebralidze et al. 2010: 60; Kakhidze& Shalikadze 2010). However, these grave assemblages have also yielded silver coinage from the times of Hadrian and Caracalla (117–217 ce; P. Jaworski pers. comm., 2018).

From the 1930s comes the accidental discovery of a richly furnished grave located in the hills near the fort of Apsaros. The skeleton was said to have been lying on a wooden bed, surrounded by numerous grave goods, which, witnesses remember, included an iron dagger, golden appliqués (for clothing?), golden bells, gilded beads, gold vessels, as well as elements of a belt of gold with terminals in the form of snakeheads and a medallion with a representation of a ‘goddess’ encircled by precious stones. According to another account, the finds included gold and silver jewellery, and at least one intaglio (Braund 1994, 185–6). It is a pity that the find as a whole was lost in unexplained circumstances. According to Georgian scholars, the artifacts resembled a gold hoard discovered at Gonio (Lordkipanidze et al. 1980, 33; see below).

The region around Apsaros is rich in hoards of coins and valuables, partly coinciding with the cemeteries discussed above. Significantly, the objects in some of these hoards, like the complete vessels and diadem from Kapandibi, suggest that they were hidden by the first owner rather than being a random set collected as investment capital. The hoards from Kapandibi, Zanakidzeebi, Makho and the Khalvachauri area seem to belong to one chronological horizon, dated by coins as well as other items to the first century bce–firstst century ce, with a strong indication to the firstst century ce. The single rich burial from the hills near Apsaros could be of a similar date.

The hoard found outside the cemetery at Kapandibi contained silver drinking vessels (a kantharos and a goblet/scythos), set with precious stones and decorated with enamel, dated to the first century ce (A. Kakhidze et al. 2015, 60–61). Silver cups in elite burials from the central European Barbaricum are usually dated to the second half of the first century ce (see e.g. Schuster 2010, 335). Other ornaments from this hoard confirm a first-century ce date (A.Kakhidze et al. 2015, 53–9; P. Gołyźniak pers. comm., November 2020). The hoard also included golden leaf-like elements from a wreath and a set of one larger (rectangular) and six smaller (lunula-shaped) pendants set with precious stones, resembling a presumed second-century diadem from the Kldeeti (Imereti) necropolis⁵ (Gamkrelidze 2014, 49). A diadem was also part of the hoard from Zanakidzeebi (A. Kakhidze et al. 1999; 2015, 94; cf. E. Kakhidze & Mamuladze 2014, 178). It was made of thin sheet gold and embossed with a schematic, linear pattern that can be dated to the first century ce (P. Gołyźniak pers. comm., November 2020). The dating is confirmed by a glass balsamarium found together with the adornment (E. Kakhidze & Mamuladze 2014, 178), and considered typical of first-century ce glasses (Isings 1957, type 28a; De Tommaso 1990, type 28; M. Wagner, pers. comm., November 2020). The jewellery, which has been preserved from the hoard found outside the cemetery in Makho, is dated to the first–second centuries ce (Braund 1994, 186; A. Kakhidze et al. 2015, 90–93; cf. E. Kakhidze & Mamuladze 2014, 178), with the first century being the more likely date (P. Gołyźniak pers. comm., November 2020). Finally, there is an amethyst gemstone, presumably from a finger ring, with an unidentified motif engraved on it, that was found in unexplained circumstances near Khalvachauri. A carnelian intaglio with an engraved shrimp image came from the same context. The two objects are dated to the late first century ce (Braund 1994, 186).

The unique collection of valuables referred to in the literature as the Gonio hoard is completely different (Lordkipanidze et al. 1980). It is usually dated to the first–second centuries ce (A. Kakhidze et al. 2015, 63–89) and although it was found at some distance from the Apsaros fort, it may nevertheless be related to the two hoards of coins that were hidden outside the fortress walls. The first of these, discovered in 1998 (Varshalomidze 2009, 28–47, 87–88, nos 63–103), contains 42 silver coins, mainly from the mints in Rome and Caesarea in Cappadocia and dated between the first century ce and the early third (terminus post quem of the deposit = 217 ce). The second of the hoards was discovered in 2018 and contains eight silver coins minted in Caesarea in Cappadocia and six large bronzes from Trebizond (terminus post quem = 235 ce; Jaworski et al. 2021, 300–301). The deposits containing silver and bronze coins were almost certainly hidden during the invasion of the barbarian Borans, which took place in the mid-third century (Kakhidze 2009: 313; P. Jaworski pers. comm., 2020), probably in 257 or 258 ce (Myzgin & Dydenko 2021). Thus, this group of finds predates the burials in the Makho and Kapandibi cemeteries, which represent the late third- and early fourth-century ce horizon.

Mapping these finds shows some interesting regularities. The sites which have yieled hoards of precious items and coins form two distinct groups. One is located around the fort at Apsaros and includes the puzzling single burial in the hills and the late hoard from Gonio. The other occupies a microregion on the banks of the Çoruh river, extending to Makho in the south and Khalvachauri in the north (Fig. 2: 2). This cluster contains both first-century ce hoards and third-century ce cemeteries. The coincidence of two different chronological horizons is at once intriguing and difficult to interpret in the absence of corresponding archaeologically verified settlement remains. However, a look at regional topography reveals that the villages of Makho and Kapandibi lie on a small spit of land in what is otherwise a lowland, once swampy, where the river exits from a mountain gorge. Not only was this a good place for settlement, but it was most probably also situated on an easily secured and convenient (perhaps even the only) natural ford in the lower reaches of the Çoruh river. More importantly, this ford seems to have been the only way to cross the river for anyone travelling the coast from north or south and also down the river from the mountains in the east.

This particular location explains the wealth of the local community, evidenced in the hoards and grave equipment discussed above. It does not, however, explain the presence of exceptional wares, such as the silver vessels from Kapandibi, customarily used for drinking wine. Tacitus mentions silver vessels as diplomatic gifts given to barbarian chieftains at the beginning of our era (Tac., Germ. 5). Similar gifts are attested in an archaeological context in the European Barbaricum in what are considered to be princely graves of the Lubieszewo horizon, dated to the first century ce (Kolendo 2017: 115; Wielowiejski 1989), more probably the second half (Schuster 2010, 335). Remains of silver vessels in grave contexts are also encountered in western Georgia (A. Plontke-Lüning, pers. comm. June 2019).

The golden diadems from Kapandibi and Zanakidzeebi are equally unique. They could have been worn by a member of the local elite (a tribal leader, perhaps) and they could even have belonged to the same community, if not the same family. Local rulers, upon being recognized by Rome (Arr., Peripl. 11), could have received special insignia (and the right to wear them) in confirmation of their new status. Phlegon of Tralles, for example, recounts the king of Bosporus Cotys being granted by Caesar the right to wear a diadem (Phlegon, Mir. 15.22). Thus, the gold diadems from Kapandibi and Zanakidzeebi could have had symbolic significance even greater than their purely material value. A ring with an amethyst gemstone like that from Khalvachauri must have also been considered special; in Ptolemaic Egypt such a ring (but with the image of the ruler) would certainly have been considered a diplomatic gift (Plut., Luc. 3.1). In any case, these assemblages, with their probable diplomatic overtones and ties with the local elite, fall within a very narrow time horizon, unprecedented in the local material culture from earlier periods and finding no continuation over time.

The question that is of particular interest is whether this flourishing settlement on the lower Çoruh river could reflect the southward migration of a group of barbarian pirates. The location was convenient, on the crossroads of important land routes, as well as on the borderland between Polemon's Pontus annexed by Rome and the tribal territories dependent on the kings of Iberia. A local ruler controlling such a position must have been a target for Roman diplomatic efforts. Close relations and the stability of the alliance were certainly strengthened by the construction (already during Nero's reign) of a fort at the mouth of the Apsaros river (Figs 2: 1 & 4) (Karasiewicz-Szczypiorski 2016; 2018; Karasiewicz-Szczypiorski & Mamuladze 2019). When the garrison was forcibly evacuated (in 69 ce?), the community stayed.

The situation changed during the reign of the first rulers of the Antonine dynasty, as recent archaeological research has demonstrated (Fig. 5). The Apsaros fort on the Çoruh river, on the border with the Kingdom of Iberia, was rebuilt in Trajan's time and then expanded (in the reign of Hadrian), unlike the presumed local settlement and power centre on both sides of the river a few kilometres away. With the border set on the river, the local settlement could have moved deeper into Roman-controlled territory, as on the Rhine during the same period. In Caesar's time, for example, Mesapians were living on both banks of the river (Caes., BGall. 4.4). The Ubi originally inhabited the eastern (barbarian) bank (Caes., BGall. 4.16 and 6.9–10). A little later, however, they were settled on the Roman side of the river as allies (Tac., Germ. 28; Plin., HN 4.31.106). A similar displacement occurred in the case of the Batavians (Tac., Germ. 29). The strengthening of the border could have been linked not only with the evacuation of their allies but also with the creation of a strip of uninhabited land in the pre-frontier (Tac., Ann. 13.54; 13.56; Cass. Dio 71.15; 72.2.4; Potter 1992). Interestingly, it was during the reigns of Trajan and Hadrian that we hear of a ruler of the Heniochi and Macrones, king Anchialus, who had his residence on the Prytanis river west of Apsaros, and was allied with Rome. A certain regularity may be noted based on this juxtaposition of facts and the information contained in Arrian's account. Before Hadrian's reign (although it is difficult to determine the exact dates), the forts of Hyssou Limen (in the land of the Sanni) and Athenai operated in the area east of Trebizond. At Apsaros, the army infrastructure from Nero's times was falling into ever deeper ruin, but an unknown local community allied with Rome lived there. During Hadrian's time, there was still a garrison in the land of the Sanni, a tribe in conflict with Trebizond and causing problems for the governor of the province (Fig. 5). However, the next garrison towards the east was located on the Apsaros, but Arrian does not mention any local community in the immediate area. There are also no archaeological traces of such a settlement. Anchialus ruled his united tribes from a residence on the Prytanis river, while the abandoned fort of Athenai was located on his lands.

The local community from the banks of the Apsaros may have relocated to lands lying just above the Prytanis (most likely due to the reconstruction of the Apsaros fort and the strengthening of the border). It is also probable that the renewed fortress at Apsaros was manned in part by soldiers withdrawn from Athenai.

The resettlement of the crossroads on the Çoruh river in the second century ce coincided most probably with the abandonment of the fort in Asparos, possibly in connection with the Boran incursion (E. Kakhidze 2008, 313, traces of destruction; Jaworski 2021).⁶ Therefore, at the time of the maximum strengthening of Roman military presence in the area, there was probably no civilian settlement in this section of the borderland.

In light of this discussion, it can be assumed that the economically and strategically important Apsaros river crossing was under the control of either the Roman garrison or a local community with close ties to Rome. However, these two models of border control were probably mutually exclusive. The strengthening of military presence at Apsaros in the second century and the first half of the third corresponded to a period for which there is (as yet) no clear evidence of civil settlement along the lower reaches of the river nearby.

Summary

Collecting and analysing scattered information on the local tribes inhabiting the area at the mouth of the Apsaros river in the first centuries of the common era has given a better understanding of the local society, which appears to have been composed of the indigenous Macrones and newcomers from the southwestern Caucasus, the Heniochi.

The migration of the latter southwards at the turn of the era may be linked to the decline of port cities and emporia on the Colchis coast (Fig. 6), possibly due to a rising wave of piracy (by the Heniochi, the Achaeans, and others). The bandits sailed the coastal waters and navigable rivers in small wooden boats called kamarai/camarae. This practice came to an end with the annexation of Polemon's Pontus by Rome and the subsequent reorganization of the Roman fleet. The construction of the first forts certainly contributed to the maintenance of an imposed order (Fig. 4). The fortifications seem to have been located in the tribal territories of the Heniochi, that is, at the mouth of the Çoruh river (Apsaros) and at the foot of the Caucasus (Sebastopolis). The last (short-lived) return to sea piracy in the region took place during the anti-Roman revolt of Anicetus (in 68–69 ce). The two garrisons were presumably evacuated at this time.

A short-lived power centre appeared in the lower reaches of the Apsaros river at the turn of the common era. The settlement could have been established by newcomers who were subsequently perhaps resettled elsewhere. Tying in this evidence with the information from written sources, one has to consider the possibility that these newcomers were part of the Heniochi group of tribes. That these people were in good relations with Rome is suggested by the silver vessels and gold head adornments that eventually found their way into hoards abandoned by the departing community. These could have been diplomatic gifts or else status-related insignia emphasizing official Roman acceptance of the power exercised by a local chieftain/king.

The apparent disappearance of the settlement cluster in the lower reaches of the Apsaros river appears to coincide with the rebuilding (and subsequent expansion) of the nearby fort during the reigns of Trajan and Hadrian (Fig. 5) when the border with the kingdom of Iberia was reinforced. This local power centre seems to have remained within Roman territory, but was moved to the mouth of the Prytanis river located a little further west. King Anchialus of the Heniochi and Macrones, a Roman ally, had his residence there. Given the described coincidence, a quest for continuity between the two power centres is justified, just as the gap in settlement at the crossroads of overland routes and the river ford during the period of the maximum reinforcement of the fort, that is, during the reign of the Antonines (and Severans), is very probable. A revival took place most probably only after the evacuation of the garrison in the mid-third century ce.

In conclusion, the model of relations between Rome and the local tribes at Apsaros, which is drawn here based on a detailed analysis of the written sources and available archaeological evidence from recent excavations, concerns tribes that inhabited the Roman province of Cappadocia, but were still ruled by their kings. Applied to other Roman provinces, this model may yet contribute to a better understanding of relations of this kind elsewhere in the Roman Empire.