Rome's power in Anatolia rested less on the size of its armies than on how quickly and regularly those armies could move. Roads, milestones, bridges and way stations together formed a system of circulation that tied the provinces to one another and the frontier zones to the interior. Amaseia, in inland Pontus, stood in the gorge where the Yeşilırmak (the ancient Iris) cuts through the mountains, and it was one of the important nodes of this system in northern Anatolia. The Amasya Governorate's account of the Roman period makes the same point: the city, a provincial centre, also became the hub of the inter-provincial road system, and the roads of Galatia and Cappadocia, for example, ended at Amasya. The roads of the Pontic kings and the city's coinage are the subject of separate articles; here they are mentioned only as far as needed to understand the Roman network.

Routes Shaped by Geography

The reason the roads follow the lines they do becomes clear from the landforms of Pontus. The territory of the old kingdom, lying between the Black Sea coast and the northern edge of the Anatolian plateau, consists of ranges and valleys running parallel to the shore. Arthur Munro's classic study of 1901 likens this landscape to gigantic entrenchments scored along a hillside: mountains and valleys follow one another from east to west, while the Iris and the Halys (Kızılırmak) pierce these ridges northwards only through narrow rifts. The northernmost chain drops steeply to the sea and cuts the coastal strip off from the interior, and the roads that cross this barrier are few and difficult. Along the coast from Amastris in the west to Trapezus in the east, Amisos is the only wide door into the interior. Between the Paphlagonian mountains and the Paryadres range the Iris and the Halys reach the sea; in between lies a slight dip, which Munro says cannot be called a gap, and the great road reaches Amasia and then Cappadocia across it. The road is not considered hard going, but it is toilsome, because the ridges and valleys continue without interruption across the hollow.

This geography also set the limits of Roman engineering. In a thesis prepared at METU, the map of Roman roads that David French derived from milestones was overlaid on digital elevation data, and the roads were found to follow valley floors for the most part. Roman roads were less straight lines defying the mountains than structures that took the natural corridors and then organised, strengthened and marked them. For this reason too, Amaseia's position depended on the gorge of the Yeşilırmak: it lay in a valley running north to south, at the natural mouth of the passage from the sea to the interior. Strabo's description of the city as both a town and a fortress, a high and precipitous rock descending abruptly to the river, shows that the same passage was also favourable for defence.

Roads Leaving Amaseia

Scholars working on the roads of Pontus speak of several main lines that complemented one another. One of them is the trade road running from Amisos on the coast down to Zela, which formed the north-south axis of the region. According to Munro this was the most important north-south link in the country, but it was not the country's principal route of internal communication; it served mainly external trade and contact with the coast. The real backbone of internal communication was the east-west trunk road running from the Lycus valley to the valley of the Amnias. Amaseia lay where these two axes crossed. According to John Wilson's study, as reported in the METU thesis, when the road reached Amaseia many routes from the north-east, north-west, west, south-west and east converged there, and the city served as the gateway through which interior Pontus reached the sea trade.

In the north the link to Amisos kept its importance in Roman times as well. A traveller setting out from the coast reached Neoklaudiopolis (modern Vezirköprü), which lay on the great road from Nicomedeia in Bithynia to Satala, and from there arrived at Amaseia. One of the older measurements of the distance is the figure Munro gives on the authority of J. G. C. Anderson: about 71.5 miles by road from Samsun to Amasia, and 99.5 miles in all to Zela. Munro uses this to show that no north-south road in Pontus could be much longer, whereas the east-west line from Niksar to the Halys alone measured 105 miles. The north-south road was short and the east-west road long; the real length of the kingdom lay along its horizontal axis.

To the east, the link with Neocaesarea (Niksar) led to an important trunk road following the Lycus valley. This road stayed on the valley floor, opened onto the fertile plain of Phanaroia, and joined an easy line running south to Comana Pontica. Comana is accepted as having been linked directly with Sebasteia, but there is no firm evidence for a road between Comana and Sebasteia; there is only a vague identification of places called Verisa and Bolus. That a road existed between Amaseia and Comana is recorded in Ramsay's work. A line from Tavium to Comana Pontica and Neocaesarea also passed through Zela, and Zela lay on the trade road coming from Amaseia. These temple cities were important nodes both religiously and economically, and Amaseia was in constant contact with the sacred and commercial centres around it.

Routes Reorganised under Roman Rule

The origins of the roads in the region predate Rome. In the time of the Pontic kings there were already routes linking the coast with the interior, guarded by fortresses and outposts. Rome did not build these from scratch but reorganised the existing lines to suit its own administrative and military needs. Munro's study notes that when Pompey organised the conquered territory he placed five of his cities on the trunk road. A remark about Zela shows the local consequences of these arrangements: under Roman rule the trade road coming from Amaseia was redirected so that it passed through the area of modern Tokat, which led to the revival of Gaziura (Turhal), a medium-sized settlement, in Roman times. A change in the course of a road could change the fate of the settlements along it.

The political framework also affected road policy. According to the Governorate, when Antony set up buffer kingdoms against the Parthian threat, the Mithridatic district that included Amasya was given in 39 BC to Darius, son of Pharnaces II. At this time Amasya was the metropolis of the Mithridates Galaticus district. When Augustus created the province of Galatia in 25 BC, this district was attached to it. For the road network this meant that Amaseia in the north now stood in the same administrative framework as the roads of the Anatolian interior and Cappadocia. According to Munro's note, Ptolemy counts this road within Pontus Galaticus, and the likely reason was to give the province of Galatia its own free access to the sea. According to the information compiled by Livius, Amaseia, which had been the administrative centre of Bithynia-Pontus, was transferred to Galatia around AD 2-3 and to Cappadocia around AD 112, and received the title of metropolis in the second century. After Diocletian it became the religious and administrative centre of Diospontus.

Milestones

The best-known markers of Roman roads are the milestones set up at regular intervals. A Roman mile corresponds to roughly 1,480 metres; the stone sometimes records the distance and always names the emperor who built, repaired or re-erected the road, with his titles. Each milestone is therefore both a road marker and a proclamation of authority. For researchers it is a document showing in which period, and under which emperor, the road was built or maintained, since most main roads are described in detail in no written source.

According to Munro's notes, Roman milestones belonging to the Neocaesarea-Amasia road have been found at Niksar, Fidi and Çalgara. Fidi is identified with Ptolemy's Pida and also appears in the Tabula Peutingeriana. On the northern branch, around Neoklaudiopolis, some thirty milestones are known; an article on a stone photographed at Havza gives the number for this area as thirty-nine, and the difference evidently depends on which study counted them. The series Anderson discovered in the summer of 1899 between the Halys and Neocaesarea shows that successive emperors, from Nerva to Constantine, had the road rebuilt or repaired; Anderson bases his view of the road's importance in Roman times on this complete series. A milestone photographed at Havza and now lost bears the name of C. Aradius Rufinus, a governor of Galatia in the third century, and documents this governor for the first time.

Systematic work has also concentrated on these stones. David French's interim catalogue of 1988 records all the milestones known up to that year with their publication histories, and later the Danish scholar Tønnes Bekker-Nielsen's research around Neoklaudiopolis added new stones. What remains of the road itself is scanty. Munro and Anderson wrote that on a stretch of old road running parallel to the nineteenth-century highway between Havza and Kavak the pavement stood intact in places. Near Aşağı Narlı, Anderson saw paving remains that might belong to the old road.

The Tabula Peutingeriana and Road Lists

Roman roads can be followed not only through archaeological finds but also through written road lists and maps. The Itinerarium Antonini and the Tabula Peutingeriana are among the foremost documents of this kind, listing stations and the distances between them. The Tabula Peutingeriana is a medieval copy of a map of the Roman road network, drawn on a long, narrow strip. The explanatory text of the scholarly atlas for Cappadocia and Armenia Minor records that the main routes leaving Caesarea and Sebasteia, and their intermediate stations, are known from these two sources, and that some segments of the Tabula reappear in the geography of the Ravenna Cosmographer, which points to a common source. The same text notes that the region's most important military roads ran from Ancyra through Sebasteia to Satala and through Caesarea to Melitene on the Euphrates; Strabo, who wrote in Amaseia and knew the region at first hand, stands out as a source.

According to the study that published the milestone from Havza, the road from Havza to Neoklaudiopolis was frequently used by Roman troops going to the eastern frontier and is mentioned in the Tabula Peutingeriana. Bekker-Nielsen, on the other hand, reminds us that the east-west route running the length of Pontus, which British travellers compared with the Grand Trunk Road in India, is described in no ancient source and does not appear in the surviving itineraries. The milestones show conclusively that it was an important Roman road, but in his view the empire's real east-west highway was the southern route through Ankara. One should therefore be cautious about the roads around Amaseia: some lines are known from written documents, others only from milestones.

Bridges over the Yeşilırmak

The most tangible remains of the road network are the bridges. The Yeşilırmak divided ancient Amaseia in two; the rock of the citadel stood on one bank of the river, and the rest of the city and its outer quarters on the other. Strabo describes the situation thus: two bridges have been built over the river, one leading from the city to the suburbs and the other from the suburbs to the outside territory, since the mountain on which the rock rests ends where the second bridge stands. Strabo then describes the valley that begins at the river, at first narrow and then widening into the plain called Chiliocomum, "a thousand villages", followed by the fertile lands of Diacopene and Pimolisene, all extending to the Halys. This description shows that the bridges were indispensable both for the interior of the city and for its connection with its agricultural surroundings.

The oldest of these bridges to survive is the Alçak Bridge. The Amasya Governorate records that it was built over the Yeşilırmak in Roman times to give passage from the ancient Amasya citadel to the quarters on the opposite bank; the local press calls it about eighteen hundred years old. As the riverbed rose over time the arches were left low, and in 1865 the governor of Amasya, Ziya Paşa, had piers built on top of the arches, keeping the bridge in use. The structure seen today is therefore a bridge in which Roman arches and nineteenth-century additions are joined together. Archaeological assessments of Amasya note that the topography of the city changed little through the Hellenistic and Roman periods, Late Antiquity and the Middle Ages, while the number of bridges linking the two banks of the river grew over time.

Beside the Alçak Bridge, the other historic bridges over the Yeşilırmak belong to the city's later ages. The Çağlayan Bridge, built in 1076 by İltekin Gazi, one of the Danishmend emirs, rests on six round arches, all in cut stone. The Künç Bridge was built by Hundi Hatun, daughter of the Seljuk ruler Sultan Mesud, and is notable for its wide arch span. The Meydan or İstasyon Bridge, a five-arched structure, was built in 1145 by the Seljuk Sultan Mesud I; when it collapsed in the earthquake of 1374 it was rebuilt in the time of Şadgeldi Paşa with its original character preserved, and in 1828 it was repaired by the Amasya notable Özlü Yusuf Ağa. The local researcher Hüseyin Menç, who says that four of the eleven bridges in the city centre are historic, notes that these four date from the Roman, Danishmend, Seljuk and Ottoman periods; the cutwaters on the upstream faces of the piers are special arrangements made to break the force of the water. The same crossing, the same link between the same two banks, has been maintained for eighteen hundred years by successive structures.

Other Roman Bridges in the Region

The bridges of the same road system are documented elsewhere in the region too and place Amaseia's bridge engineering in a wider frame. At Comana Pontica, reused civic inscriptions have been found at the base of the Roman bridge over the Iris, and Hamilton recorded this bridge; a later regulation project, however, destroyed it. On the Kızılırmak, of the old bridge west of the ferry crossing at Çeltik only a fragment of a pier on the west bank and the great abutment on the east bank remain. The structure Anderson described had a core of small stones and a casing of large blocks fitted without mortar, and must have crossed the river in a single arch of more than a hundred feet. Bekker-Nielsen, judging from the construction technique, says the bridge may be Roman, and that a span of more than a hundred feet, though not usual for Roman bridges, would not be unique. The bridge now lies beneath the waters of the Altınkaya Dam. The Kurt Bridge near Vezirköprü is dated to the Seljuk period, but the possibility that a Seljuk or Ottoman superstructure rests on Roman foundations can only be tested by excavation.

Trade, Army and Communication

The most visible function of Roman roads was military. The lines running from the coast to Amaseia and on to Zela and Cappadocia were the corridors through which troops and supplies bound for the eastern frontier passed. The same roads were indispensable for trade. Amaseia lay on the trade road linking the port of Amisos on the coast with the Anatolian interior, and the fertile plains around it fed the city's wealth. This road network did not merely carry goods; officials, soldiers, merchants and the cultures of east and west also reached the city along the same routes.

One sign of commercial vitality is the city's own coinage. The Amasya Governorate states that civic coins were struck in the city from Domitian (81-96) to Severus Alexander (222-235) and that many of them are in the coin collection of the Amasya Museum. An earlier assessment also mentioned bronze Amaseia coins from the reign of Caracalla. It has been argued that under Mithridates VI the cities of interior Pontus struck coins of similar types, that these circulated only within the kingdom, and that this points to regional economic integration; the details of those coins are a matter for another article. That the city struck its own money in Roman times reflects the economic weight and administrative standing it had gained through the road network.

Continuity Imposed by the Valley

The present road network does not repeat all of the old routes, but the modern roads that follow the Yeşilırmak valley and the Roman routes are products of the same topography. The fact that Roman roads followed valley floors explains why settlements remained tied to the same corridors. Amaseia's position between the sea and the Anatolian interior was its defining feature before and after Rome; Roman roads placed this position in a new order and made it visible with milestones and bridges.

Chronology

39 BC: Antony, in his arrangement of buffer kingdoms against the Parthian threat, gives the Mithridatic district that includes Amasya to Darius, son of Pharnaces II.

25 BC: Augustus creates the province of Galatia; the Mithridates Galaticus district is attached to it.

AD 81-96: Civic coins are struck at Amasya from the time of Domitian; the series continues until the time of Severus Alexander (222-235).

About AD 112: According to Livius, Amaseia is attached to the province of Cappadocia.

1076: İltekin Gazi has the Çağlayan Bridge built.

1145: The Seljuk Sultan Mesud I has the Meydan (İstasyon) Bridge built.

1374: An earthquake destroys the Meydan Bridge; it is later rebuilt.

1828: The Amasya notable Özlü Yusuf Ağa has the Meydan Bridge repaired.

1865: The governor of Amasya, Ziya Paşa, has piers built on the arches of the Alçak Bridge.

1899: J. G. C. Anderson discovers the series of milestones running from Nerva to Constantine between the Halys and Neocaesarea.

1901: Arthur Munro publishes his article on the royal and Roman roads of Pontus.

Bibliography

Primary sources

  • Strabo, Geography XII.3.39 (H. L. Jones translation at Livius.org).

Academic studies

  • J. A. R. Munro, "Roads in Pontus, Royal and Roman", The Journal of Hellenic Studies 21, 1901, pp. 52-66.
  • METU graduate thesis on the temple-states of Comana Pontica and Zela and their place in the Roman road network (etd.lib.metu.edu.tr/upload/12606089).

Books

  • Barrington Atlas, explanatory text for the Cappadocia and Armenia Minor maps (Princeton University Press). assets.press.princeton.edu

Other sources

  • Republic of Türkiye Amasya Governorate, "Roma Dönemi" and "Köprüler" pages. amasya.gov.tr
  • Tønnes Bekker-Nielsen, "350 Years of Research on Neoklaudiopolis (Vezirköprü)".
  • "Havza'dan Yeni Bir Miltaşı Üzerinde Aradii Soyundan Yeni Bir Galatia Valisi: C. Aradius Rufinus", Philia.
  • "Öntarih ve Klasik Dönemlerde Amasya: Güncel Arkeolojik Bulgular", TÜBA-AR. amasya.gov.tr

Online sources