- 1 E.g. Lund 2015; Lysandrou, Agapiou 2016; Kyriakou 2019; Vionis, Papantoniou 2019; Deligiannakis 202 (...)
- 2 LAF is funded by the Carlsberg Foundation and hosted by the Saxo Institute, University of Copenhage (...)
1In 2004, Tønnes Bekker-Nielsen published his study of the Roman roads combining the study of maps, written sources and ground observations. With the advent of digital Geographic Information Systems, this road network (fig. 1) has become a standard background to many studies of the Cypriot landscape whose findings by and large confirm the reconstruction of the island’s highway network.1 However, no historical or archaeological study is ever complete, and the growing body of work motivates a revisit. This article presents the findings of the fieldwork carried out by Tønnes Bekker-Nielsen, Kristina Winther-Jacobsen and Marit Jensen in November 2023 as part of the field campaign of “Life at the Furnace: Copper producing settlement ecosystems in the north-western Troodos” (LAF).2
Figure 1 — The road network of late Roman Cyprus, after Bekker-Nielsen 2004, with subsequent emendations.
1) revised course of road in the Karkotis Valley; 2) revised course of road from Salamis to Aphrodision; 3) newly discovered road in the Peristerona valley; 4) hypothetical northward continuation.
1) Winther-Jacobsen 2013; 2) Bekker-Nielsen 2014; Bekker-Nielsen 2016; 4) T. Bekker Nielsen © LAF.
- 3 Given, Knapp 2003; Given et al. 2013a; Given et al. 2013b.
- 4 Moore 2003, p. 282; Winther-Jacobsen 2022, p. 377.
- 5 Hadjichristofi 2023. See also this volume, pp. 33‑46.
- 6 Moore 2003, p. 282.
- 7 Winther-Jacobsen 2022.
2The multi-disciplinary landscape project working in the central northern Cyprus – LAF – has three objectives: to map the relationship between mining, smelting, forestry and settlement decision-making and its sustainability; to place the copper production of Cyprus within the economic network of the Roman Empire; and to reconstruct the Late Holocene landscapes of the research area. Since the early 1990s the northern Troodos has been the focus of two large scale multidisciplinary landscape projects. LAF stands on the shoulders of the Sydney Cyprus Survey Project (SCSP) and the Troodos Archaeological and Environmental Survey Project (TAESP) building on the data and results of these projects.3 SCSP and TAESP have surveyed a large part of the area of the major copper deposits in the central northern Troodos between two of the Iron Age kingdom city centres at Tamassos and Soloi – as far as possible in the case of Soloi (into the buffer zone), which is located in the occupied zone of Cyprus. The landscape of central northern Cyprus has revealed many production sites, many of which are connected directly or indirectly to the extraction of copper. During the Roman period, Tamassos appears to have been reduced to a middle ranking settlement, probably a market town, and only Soloi qualifies as a Central Place.4 However, the dearth of middle ranking settlements is remarkable. The recently excavated Roman villa at Akaki is evidence of a type of settlement rarely identified in the area introducing a new type into the settlement hierarchy,5 although small estates and villas were suggested as an explanation for the diffusion of finds surrounding Tamassos.6 Imported table wares penetrated deep into the hinterland, which suggests the existence of middle rank settlements redistributing foodstuff intra-regionally as well as goods from the Central Place to the hinterland production sites.7 Reconstruction of the inter-regional road system provides an important insight into this economic network.
- 8 A preliminary report on the fieldwork of the “Life at the Furnace” project in 2020 and 2023 is fort (...)
- 9 Winther-Jacobsen 2013.
3At the heart of LAF is the copper smelting site, Agia Marina Xyliatou-Mavrovouni in the Peristerona valley between the SCSP and TAESP areas. The settlement ecological approach adopted by LAF provides the impetus to integrate this production site in the settlement hierarchy of central northern Cyprus, in the pan-island infrastructure and in the wider Roman economic network.8 The route of the trans-island highway between Tamassos and Soloi was resurveyed with special attention to the section crossing the Peristerona valley as well as a branch road running parallel to the Peristerona river. In addition, the fieldwork also reevaluated the evidence for a route along the Karkotis valley, an area west of the Peristerona valley studied in detail by the TAESP and published after the above-mentioned study by Bekker-Nielsen.9
- 10 Bekker-Nielsen 2000, p. 202; Bekker-Nielsen 2001, p. 253.
4For the larger part of the late Archaic and Classical periods, Cyprus was under Persian domination, with its cities and their kings enjoying some measure of self-government, including the right to strike their own coinage. Alexander’s conquests and the subsequent breakup of his empire transformed the political geography of the island. During the Wars of the Successors, the kingdoms were suppressed and Ptolemy I of Egypt emerged as master of the island, which from ca 295 BC onwards formed an Egyptian province. Amathous may originally have been envisaged as the main Egyptian base on the island,10 but in the event, the Egyptian governors established themselves at Nea Paphos, which remained the administrative capital of Cyprus until the late 3rd or early 4th century AD, when the Roman governor’s residence was transferred to Salamis, probably in connection with Diocletian’s reorganization of the provinces.
- 11 Bekker-Nielsen 1993, pp. 188‑189.
- 12 Bekker-Nielsen 2004, pp. 104‑108.
5Many of the “Roman” roads of Cyprus no doubt follow older routes laid out by the rulers of individual cities on a dendritic or “solar” pattern, radiating from each city into its periphery,11 which were later integrated into an island-wide highway system by the Ptolemies or the Romans.12 In the process, some roads were realigned or reengineered to Roman standards, others improved by the construction of switchback ramps or shortcuts.
- 13 Kassianidou 2011; Kassianidou 2013a; Kassianidou 2013b; Kassianidou 2022.
6Since the Bronze Age, the rich copper deposits have been an important resource for the island of Cyprus. The deposits are associated with the pillow lava exposed in the foothills around the Troodos mountains.13 By the Hellenistic and Roman period, the major urban centres are located near the coast and never in the immediate vicinity of ore deposits and consequently, the issue of territorial control and infrastructure must always have been important. To produce copper, the ore had to be melted in a process, which required large amounts of charcoal, and the work required labour, which had to be fed.
- 14 Bruce 1937, p. 642.
- 15 Manning 2013, p. 49, p. 54.
- 16 Urwin 2013.
- 17 Westholm 1936, p. 407.
- 18 Mitford 1980, p. 1331; see also Moore 2003, p. 282.
7The production of copper fluctuated over time. Skouriotissa, the largest slag heap in Cyprus, is estimated at 2,000,000 tons of slag.14 Radiocarbon and dendrochronological data suggest that the accessible part of the slag heap was built up from the early‑to-mid 4th to the mid‑to-late 7th century.15 Neil Urwin has calculated that “the build‑up of the slag heap at Skouriotissa reflects a daily need for 18 tons of charcoal from an area of 516 km2 within the lower slopes of the Troodos mountains, which would have required 200 daily donkey trips per day”.16 At Skouriotissa, the raw material was processed close to the ore deposit, but afterwards the copper ingots needed to be transported 9 km to the nearest harbour at Soloi, and after this harbour silted up in the 4th and 5th century AD, a further 6 km west to the harbour at Limenia.17 Most ore deposits were located some distance from the sea. Copper from the mines associated with the inland territory of Tamassos would have to travel 29 Roman miles (ca 43 km) overland on the trans-island highway to be shipped out at Soloi.18
- 19 Fourrier 2013; Papantoniou, Bourogiannis 2018.
- 20 Papantoniou, Bourogiannis 2018, p. 14.
- 21 Papantoniou 2013; Vionis, Papantoniou 2017; Vionis, Papantoniou 2019; Kyriakou 2019.
8During the Iron Age, numerous sanctuaries appear in Cyprus in a well-documented, diffused pattern, which reflects the socio-political system providing a mechanism for the Archaic and Classical city-kingdom authorities to organise and control their hinterlands.19 The importance of control and exploitation of agricultural lands is underscored by the proximity to agricultural land, which is characteristic of many Cypriot extra-urban sanctuaries of the Archaic and Classical period.20 Papantoniou and Bourogiannis did not view this as evidence for direct involvement of the sanctuary in economic activities, and the Iron Age sanctuaries have not been linked specifically to the road network, but it seems obvious that organization and control of the hinterland resources required access, and the “sacred landscape” provides a possible model for understanding e.g. movement, economic and ideological power in the Cypriot landscape.21
- 22 Papantoniou 2013, fig. 5; Rautman 2002; Gordon 2018.
- 23 Rautman 2004, pp. 147‑149; Keane 2021, pp. 3‑4.
- 24 Kyriakou 2019, pp. 48‑50, fig. 1.
9During the Roman period, the number of extra-urban sanctuaries declined dramatically, reflecting a highly centralized urban culture.22 However, with Christianity the location of sanctuaries again became diffused when rural churches spread during the 5th century.23 According to Niki Kyriakou, the early Christian rural basilicas are primarily located on agricultural land and next to important infrastructure, reflecting the economic and ideological power of the church.24
- 25 Kyriakou 2019, fig. 6; Vionis, Papantoniou 2019, fig. 11.3.
- 26 Given et al. 2013b.
- 27 Caraher, Moore, Pettegrew 2014; Rautman 2000.
- 28 Papacostas 2013; Papageorgiou, Bakirtzis 2018; Vionis 2019, p. 74; Winther-Jacobsen 2022, pp. 16‑17 (...)
- 29 Feissel 2013, pp. 220‑222; Nicolaou 1964, pp. 197‑198 no. 10.
- 30 Feissel 2013, pp. 227‑234.
- 31 Feissel 2013, pp. 222‑225; Keane 2024, p. 195.
10The distribution map produced by Kyriakou reveals the association between early Christian churches and rural establishments, regional centres, ports or coastal emporia as well as the infrastructure connecting these.25 The map also reveals a dearth of early Christian churches in the hinterland of Soloi during a period associated with a great increase in rural settlement and industrial activity,26 not only in this region but island-wide.27 In the Karkotis valley, where three villages have been linked to the Roman mining activity at Skouriotissa, only one of the recorded 13 churches – Panagia Skouriotissa, the church next to the Skouriotissa slagheap – can be dated earlier than the 11th century.28 In 2009, a reused inscription referring to Bishop Ioannes was found in the church of Agia Eleousa in Korakou in the Karkotis valley ca 12 km south of Soloi,29 a church dated to the 15th-16th century. Denis Feissel gives a critical reading of four inscriptions and fragments associated with Bishop Ioannes of Soloi. The inscription from Korakou relates his restoration of a dilapidated structure, but the type of building is not stated. Ioannes is famous for undertaking the rebuilding of the basilica at Soloi in 655,30 but another inscription refers to his restoration of a hostelry.31 Consequently, the inscription from Korakou provides no evidence for the presence of an early Christian church in the area, but it does demonstrate the close connection between Soloi and the hinterland.
- 32 Deligiannakis 2022, pp. 114‑117.
- 33 Papacostas 1999, vol. 3, pp. 18‑19 no. 24.
- 34 Kyriakou 2019, fig 6. Personal communication with Catherine Keane, who is currently undertaking a s (...)
11On the one hand, the evidence from Skouriotissa demonstrates that copper production did not decline during the early Christian period. On the other hand, the scarcity of early Christian churches in the hinterland of Soloi seems to suggest that the church was slower to develop as an economic and ideological power in this area. The basilicas in Soloi and Tamassos are the two earliest purpose-built Christian sanctuaries attested in Cyprus, and Deligiannakis has suggested that this is no coincidence since these two towns are also associated with some of the richest copper deposits.32 Apart from these, the only “early” Christian basilica near the important trans-island highway connecting Soloi and Tamassos is the 8th-10th century basilica of Agios Varnavas and Ilarionas in Peristerona,33 which is believed to have replaced an earlier church.34 3.6 km south of Agios Varnavas and Ilarionas is the medieval monastery of Agios Nikolaos.
- 35 Lysandrou, Agapiou 2016, fig. 2.
12The studies of Kyriakou indicate that major early Christian churches are likely to be associated with important nodes or edges in the communication network. The basilica of Agios Varnavas and Ilarionas is located more than 5 km north of the accepted route of the east-west highway, and the only evidence for an ancient settlement in Peristerona are Roman graves.35 In the absence of excavation, the method chosen to evaluate if Agios Varnavas and Ilarionas was located on the trans-island highway was to restudy the route of the road along the central northern Troodos foothills and the road through the Peristerona valley.
13An important source for the ancient road network of Cyprus is the so‑called Tabula Peutingeriana, a medieval copy of a late Roman road map (fig. 2). According to the latest study the Tabula should not be interpreted as a snapshot of the Roman road network at the time when the archetype of the present copy was created in the 4th or early 5th century AD, but rather as an accumulated mass of geographical information, parts of it going back as far as the Hellenistic period.36 On the Tabula, two major routes are indicated: a peripheral road following the coast and a trans-island highway from Soloi (Soloae) through Tamassos (Tamiso) and Tremithousa (Thremitus) with a branch to Kition (Cito). The continuation from Tremithousa to Salamis (Salamina) was included on the original but has been omitted by the copyist, though the mileage (xviii) is indicated.
Figure 2 — Cyprus as depicted on the Tabula Peutingeriana, after the facsimile edition of Miller 1962.
National Library of Australia.
- 37 Mitford 1940-1948, pp. 8‑9; for the date, Christol 1986, p. 11.
- 38 Kantor 2014, p. 261.
14The geographical compendia by Strabo of Amaseia (1st century AD) and Ptolemy of Alexandria (2nd century AD) contain a good deal of information about Cyprus, as does the Natural History of the elder Pliny (1st century AD). According to Ptolemy (5.13.5), the island was divided into four parts: the Salaminian (east), the Paphian (west), the Amathusian (south, including mount Olympos, present-day Troodos) and the Lapethian (north). This presumably reflects an administrative division going back to the period of Ptolemaic domination. It was certainly in place by the last decades BC, when the cives R(omani) Paphiae diocen(seos), “the Roman citizens of the Paphian district” dedicated a statue of the proconsul,37 Egypt itself was divided into districts (nomes) and it would be natural for the new rulers of Cyprus to organize the island along the same lines. Roman administration, by contrast, was normally based on city territories, but when the Romans found a functioning administrative scheme already in place, they might be content to let it continue, as in the case of Asia.38
- 39 Bekker-Nielsen 1995a, pp. 45‑47.
15A modern source of considerable value is the British one-inch map of Cyprus, also known as the “Kitchener map”. Immediately after the establishment of the British protectorate in 1878, H. H. Kitchener, then a captain in the Royal Engineers, was commissioned to undertake a trigonometric survey of the island which was later (1884) published as a set of fifteen sheets at a scale of one inch to one English mile (ca 1:63,300). Though the map suffers from several defects, in part due to the difficult circumstances in which it was produced, it remains useful for distinguishing between roads of modern date and those in existence before the end of Ottoman domination in Cyprus and marked on the map as “tracks”.39
- 40 Bellamy 1903, p. 14.
- 41 Ibid., p. 29, p. 32.
16More problematic are the routes marked on the Kitchener maps as “Roads constructed since English [sic] Occupation”. A comparison with C. W. Bellamy’s Monograph of the Main Roads of Cyprus (1903), which is based on the annual reports and expense statements of the public works department, is instructive. For instance, the highway from Nicosia via Peristerona to Astromeritis and Lefka is shown on the map as “constructed after the occupation”, yet according to the works department’s report for the year 1888-1889 cited by Bellamy, construction had by this time reached no further than Peristerona, beyond which point the road had only been “set out and marked with cairns”.40 Ten years later, work on the Astromeritis-Lefka section was still ongoing at the rate of two to six miles a year.41 Evidently Kitchener and his cartographic team had taken the works department’s somewhat optimistic plans at face value and assumed that all of the projected roads would be completed by the time the map was published.
17The Mesaoria is a vast alluvial plain, stretching across the island from the bay of Salamis to the bay of Morphou and bounded, as its name – literally, ‘between the mountains’ – implies, by mountains on two sides: to the north, by the Kyrenia or northern range, to the south by the Troodos. At the centre of the Mesaoria, hills are low and depressions shallow, but at its edges, runoff from the foothills has created gullies and valleys that can be wide as well as deep. The valley of the Peristerona river (fig. 3) is one such feature, that of the Potami river (fig. 5) another. The natural vegetation is sparse and mostly confined to the slopes; in the plain, farmland is only occasionally punctuated by copses or patches of maquis.
Figure 3 — Ancient and modern roads in the upper Peristerona valley; contours at 20 m intervals.
T. Bekker Nielsen © LAF.
- 42 Bekker-Nielsen 2004, p. 8.
18In general, the western Mesaoria is not an ideal environment in which to search for vestiges of ancient roads. In the plains, the loose soil is in high demand for cultivation and easily ploughed, obliterating unused tracks or roads across the fields.42 At the margins and in the foothills, the landscape has continued to evolve since the Roman period, creating new gullies or filling older ones. In the mining regions of the foothills, modern quarrying and mining using heavy machinery has transformed parts of the landscape beyond recognition. Furthermore, the region is bisected by the UN buffer zone, which passes immediately north of Peristerona, while the northern Mesaoria remains under Turkish occupation.
- 43 E.g. Bekker-Nielsen 1995b, p. 94, p. 117, p. 120, p. 122.
- 44 Camerlenghi 2018, pp. 47‑48. We are grateful to Catherine Keane for this reference.
- 45 Esch 1997, p. 20.
- 46 Lysandrou, Agapiou 2016.
19No remains of Roman bridges have been identified though the existence of several, no doubt constructed of wood, may be surmised. Slab paving is rarely found in Cyprus and the few examples defy precise dating, but in a number of places, cuttings, ledge roads, or steps carved into the rock testify to the presence of an ancient route.43 Buildings adjacent to the road-line can provide ante quem dates; if a structure is located close to the edge of a straight section of road, the road was presumably in place before the building was erected. Late Roman or early medieval churches located adjacent to an existing Roman road are a well-attested phenomenon, e.g., on the via Ostiensis44 and the via Appia45 south of Rome. Ancient necropoleis, which likewise tended to be located along roads, may also provide important pointers to the course of ancient roads.46 Since roads often also served as boundaries between fields, their alignments would sometimes be preserved in the pattern of field division, the line of a hedgerow or boundary wall, or as borders between village territories.
- 47 Bekker-Nielsen, in press.
20A shared characteristic of Roman roads in plains or foothills is a preference for straight alignments and a tendency to exploit natural gaps in the landscape. Their longitudinal profiles often reveal the care which the surveyor has taken to avoid “lost ascent”, energy expended in order to climb a gradient only to go down again a short distance later.47 A further shared feature is the deep roadbed, 1.5‑2 m deep, with a foundation of large stones. This distinguishes ancient roads from more recent ones built by British engineers using the “Macadam” method of construction with a shallow layer of crushed stone. Where the gravel surface of a disused Roman road has been destroyed by ploughing or concealed under later humus formation, its foundation layer may still give its presence away. Furthermore, since the deep stone foundation provides drainage, its course will sometimes show up as a linear parch-mark in cultivated fields.
- 48 Bekker-Nielsen 2013, pp. 12‑13.
- 49 Ohnefalsch-Richter 1913, p. 188.
- 50 Bekker-Nielsen 2013, pp. 15‑16.
21Wheeled vehicles were used in Cyprus from an early date with the two-wheeled cart – light, of simple construction and capable of turning on a tight radius – as the dominant type.48 Pack-animals would also have played a significant role, especially in hilly regions where not all roads were suitable for vehicles but could be traversed by trains of mules or donkeys. Up until the end of the nineteenth century, before the advent of road and rail transport, the mule drivers of Athienou operated a flourishing carrier network stretching across the island49 and as late as the 1990s, donkey trains could be encountered on the remoter roads of the western Troodos or the Akamas. Pack-trains supplying the animal flocks of the uplands with sea salt could carry copper, forestry products, pottery or wine on their return journey to the coast. Grain, on the other hand, was more likely to travel by cart.50
- 51 Bekker-Nielsen 2001.
- 52 Bekker-Nielsen 2004, pp. 191‑192.
- 53 Mitford 1981, p. 1331.
- 54 Papantoniou, Bourogiannis 2018.
- 55 Winther-Jacobsen 2022.
22In the early 20th century, the main port of export for Cypriot copper ore was Soloi, which was connected to the mining hinterland by a narrow-gauge railway. A similar, but shorter line connected the port of Drapia on the south coast to the mines at Kalavasos. Given the dominant position of Soloi in the early modern copper trade, and unaware of the existence of a land route from Tamassos to the south coast at Amathous/Neapolis51 and a branch from Dali to Kition,52 T. B. Mitford assumed that “for the export of her copper Tamassus must have used the port of Soli” since “Salamis and Citium were reached only through Tremithus”.53 In actual fact, Tamassos is more or less equidistant by road from the ports of Soloi, Kition, Kyrenia and Amathous, any one of which could have been used for export of copper from Tamassos. For the mines of the Peristerona and Karkotis valleys, Soloi was the nearest port, and archaeological evidence suggests that the Morphou Bay was controlled by Soloi already during the Iron Age.54 During the Roman and late antique period, the ceramic evidence clearly establishes a close connection between Soloi, the hinterland of the Peristerona, Lagoudera and Karkotis valleys and the Morphou bay.55
23In addition to these “city ports”, numerous opportunistic ports and landing-places existed on the coast of Cyprus,56 but at the present state of our knowledge, it is difficult to assess their role, if any, in the copper trade.
- 57 Bekker-Nielsen 2004, pp. 173‑175.
24As mentioned earlier, the highway leading from Tamassos across the Mesaoria to Soloi is indicated on the Tabula Peutingeriana (fig. 2), which gives the distance between its end-points as 29 Roman miles, equivalent to 43 km. Since this is only marginally longer than the distance as the crow flies (41 km), the Roman road must have followed a nearly direct course with few and minor deviations en route.57
- 58 Bekker-Nielsen 2004, p. 176.
25East of the Peristerona river, a string of hills and high ground runs parallel to the river and the modern highway from Peristerona through Orounta to Kato Moni. About 2.5 km south of the village centre of Orounta, this line of hills is broken by a gap, clearly visible on the skyline when viewed from across the valley.58 An ancient road surveyor would undoubtedly have taken advantage of this passage (fig. 3).
- 59 Price 1980, p. 94, Nicosia 16; Nicolaou 1967, no. 43.
- 60 LAF is undertaking further survey in this area in November 2024.
26The descent from the high ground to the eastern bank of the river is not immediately traceable in the field. Further down, parts of the right bank have been eroded away, exposing a section through a massive pair of stone foundations made with river cobbles and small boulders. Pottery collected from the infill between the foundations suggests a post-antique date (fig. 4). Two sites of the Early and Middle Chalcolithic periods indicated by pottery and lithic scatters located across from each other, on either side of the river at Orounta-Stavrou tou Kounti and Monarka-Dimma tis Peristerona, attest to the importance of this river crossing in the longue durée.59 The settlement at Stavrou tou Kounti however also includes a Roman phase.60 The road crossed the river in the vicinity of UTM 36S 508087/3881932, continuing in a north-northwesterly direction before swerving northwards to avoid a patch of uneven ground at UTM 36S 507810/3882275.
Figure 4 — Stone foundations of uncertain date exposed in the right (east) bank of the Peristerona river valley, UTM 36S 36S 508119/3881859.
T. Bekker Nielsen © LAF.
27The road now ascends the western slope in series of straight sections. At UTM 36S 507488/3882277 the road intersects the modern tarmac road leading from Peristerona towards Agia Marina. Beyond the intersection, the roadline is lost in the fields but reappears after a hiatus of 1.9 km at UTM 36S 505982/3883507. From here, the route can be traced as a gravelled roadway, 3‑4 m wide, for ca 3 km. It consistently keeps to the high ground on the north side of the valley, descending on a steady gradient, running in rectilinear sections, making only minor corrections of orientation to adapt to the landscape and generally avoiding lost ascent. Adjacent to the solar farm at UTM 36S 505542/3883698, it intersects a modern dirt road coming from Peristerona and continuing into the valley.
28From the point of view of the road surveyor, the most challenging section was that west of a stream crossing and the low saddle at UTM 36S 505062/3884051. Here the road swerves northeast and takes on the character of a ridgeway, with terrain falling away to both sides.
29The road continues along the high ground north‑east of Potami village (fig. 5), where a low, linear earthwork (fig. 6) can be observed, running parallel to the modern road over a distance of ca 150 m, from UTM 36S 503493/3884780 to UTM 36S 503348/3884809. The present width of the bank is ca 3 m (consistent with a Roman date), the elevation above the modern road surface is ca 1.5 m. This feature may represent the agger of a Roman road. Large stones not resembling those found in the fields to the north or south spill out of the southern face of the earthwork and on to the modern roadway; if this is indeed a Roman road bank, these stones could derive from its foundation course. This hypothesis could be tested by excavating a section across the bank.
Figure 5 — Ancient and modern roads around Potami village; contours at 20 m intervals.
T. Bekker Nielsen © LAF.
Figure 6 — Raised bank, possibly the remains of the Roman agger, to the north of the road at UTM 36S 36S 503427/3884792.
T. Bekker Nielsen © LAF.
30West of this point, the road‑line is not immediately visible in the fields but reappears at UTM 36S 502985/3884926 as a dirt road. Between UTM 36S 502868/3885004 and UTM 36S 502766/3885109 the present road makes a detour around two properties. At UTM 36S 502726/3885154, where the level of the modern road surface is ca 1.5 m lower than the field adjoining the road to the north, scatters of large stones can be seen in the slope, reminiscent of the linear earthwork east of the village. The road, which now has a hard surface, continues westward along the high ground on the north side of the valley, until the modern road descends on an S‑curve to cross the Potami river. The ancient road must likewise have crossed the stream, but its route across the valley floor and ascent on the far side can no longer be traced. It probably diverged from the modern road in the vicinity of UTM 36S 502602/3885241.
31The ancient road‑line reappears at the intersection with the Astromeritis-Vyzakia road at UTM 36S 501350/3885733. West of the intersection, the road‑line remains visible as a dirt road ca 4 m wide running on a linear alignment for ca 850 m. Apart from one point where it swerves to avoid a wooden electricity pole rather incongruously placed on the centre line of the road (fig. 7) it makes only two minor corrections of direction en route.
Figure 7 — At the intersection with the Astromeritis-Vyzakia road, the road‑line is blocked by the modern embankment (on left). UTM 36S 36S 500505/388593.
T. Bekker Nielsen © LAF.
32Ca 3 km northwest of Potami village at UTM 36S 500505/3885936, the present highway from Astromeritis towards Troodos crosses a small stream on a concrete culvert. The site is bordered on the west by the UN buffer zone less than 1 km distant and by the ongoing earthworks for the projected motorway.
33In 1963, the road‑line west of the stream was visible in the landscape and on DLS photographs of that year. It intersected the Astromeritis-Troodos highway, then merged with the line of the highway ca 150 m further west. The ancient and modern roads were presumably contiguous for the next 1.2 km before the ancient road‑line turned west-northwest towards Elea, probably along the high ground on the right bank of the Potamos tis Elias.
34At some time between 1963 and 1973 (the new alignment is shown on the DLS 1:50,000 map of 1973), the highway was rerouted on a more direct course. At the stream crossing, the roadway was now elevated on an embankment with a double concrete culvert. In the process, the junction of the field road with the highway was shifted westwards and connected to the older road‑line by an S‑shaped descent more or less parallel to the slope of the embankment. Inevitably, all traces of the older road to the west of the junction were obliterated by the earthworks for the new carriageway.
35Beyond the junction, the roadline enters the UN Buffer Zone and was not surveyed.
36A short distance beyond the river crossing south of Orounta, the modern farm track meets a dirt road leading almost directly north on a linear alignment for 1 km (fig. 3), making a slight deviation towards the river and skirting the Medieval church precinct of Agios Nikolaos where the present-day road-line is blocked by monastery buildings of 20th century date. North of the church, the alignment reappears first as a field boundary, then as a dirt road leading northwards for ca 600 m from UTM 36S 507763/3884389.
37At UTM 36S 507688/3884829 the present-day dirt road deviates westwards, running parallel to the line of a modern concrete irrigation channel. North of this point, the valley floor widens, leaving ample space for the river to meander off its bed, as it no doubt has done several times since antiquity, and forcing the present-day route onto a more westerly course. The ancient road-line presumably continued straight ahead on a line 50‑100 m further east, closer to the present bank.
38At UTM 36S 507875/ 3885875 the road‑line reappears as a modern gravel road over a distance of ca 250 m (fig. 8). The modern road then turns to pass east of a fenced precinct at UTM 36S 507886/3886257, whereas the older alignment continued ahead, its course marked by a sequence of hedgerows and field boundaries until it is lost in a recently planted olive grove at UTM 36S 507839/3886603. At the northern edge of the grove, a retaining wall may mark the line of the former road. From here, it continues as a road until UTM 36S 507573/3887034 where the modern dirt road turns left (west) while a field track continues straight ahead.
Figure 8 — 1 km south of Peristerona, the road is preserved as a gravel roadway running parallel to the river, UTM 36S 36S 507884/3886019
T. Bekker Nielsen © LAF.
- 61 Papacostas 1999, vol. 3, p. 19 no. 24.
39At the entrance to the built‑up area of Peristerona village, its course is taken up by an alley leading south to north and terminating at a group of buildings south of the church square. If the alignment were extended northwards, it would pass through the narthex of the church of Agios Varnavas and Ilarionas, a 19th century addition to the medieval structure.61 Before the construction of the narthex, the main door of the church opened directly onto the road.
40North of the church, the course of the road is not immediately traceable. It would have been forced westwards by the river and part of it may be preserved in the present-day Grigori Afxentiou street leading to the intersection with the B9 highway from Nicosia to Astromeritis. North of the highway, Agiou Georgiou street runs on a south-north course, passing close by the cemetery of Peristerona at UTM 36S 507203/3888025. Beyond this point, the alignment enters the UN Buffer Zone and a field survey was not carried out. Satellite images, however, show a line of hedgerows and field boundaries continuing the line of the road along the Peristerona valley northwards through the Buffer Zone. The alignment is nearly dead straight over a distance of 6 km and the lost ascent is negligible until the Serrakhis river is reached at UTM 36S 506688/3893305, south of Masari (Şahinler) village and ca 2 km below the confluence with the Peristerona river. This crossing and the remainder of the road are within occupied territory and were not surveyed in the field.
- 62 Shown as an overprint on the 1983 edition of the 1973 1:50K map.
41If this alignment does in fact preserve the line of the ancient road, and assuming that its ultimate destination were Lapethos, it would have continued in a north-westerly direction to pass through the northern range at Panagra (Geçitköy). The route leading northwest from Masari appears on 1963 aerial photographs as a gravel road running on a straight line and oriented on the centre of Kyra (Mevlevi) village. Between 1973 and 1983, this road was re‑engineered and resurfaced as a motor road.62 From Kyra, a modern road continues north-west, hugging the foothills, to merge with the highway from Morphou (Güzelyurt) shortly before reaching Kalokhorio/Kapouti (Kalkanlı) while another road follows a slightly longer trajectory across the valley floor. One of these may represent the ancient road‑line.
- 63 Koucky, Steinberg 1982, p. 130. Koucky and Steinberg call the Agia Marina Xyliatou-Mavrovouni slagh (...)
- 64 Winther-Jacobsen et al. forthcoming.
42At its southern end, the centre of the “Life at the Furnace” project – the slagheap at Agia Marina Xyliatou-Mavrovouni – is located ca 2.5 km south of the river crossing at UTM 36S 508087/3881932, on a small plateau on the west side of the river at UTM 36S 507150/3879789. Koucky and Steinberg dated this slagheap to the Medieval period.63 However, the project dates the slagheap to the Late Roman period based on the type of slag cakes and slagged pottery found in the slagheap.64 Access to the slagheap from the west or east is difficult due to the nature of the terrain, whereas access would have been easy from the north, following the river valley, and presumably the copper travelled out along this route, through a narrow pass in the foothills.
43One may assume that a road, probably not an engineered roadway passable for vehicles but rather an informal route for pack-animals, connected this smelting site with the east-west road and the road leading north to Peristerona and beyond. If so, the route presumably followed the left bank in order to avoid crossing the river, keeping to the high ground as far as the present-day dairy farm at UTM 36S 507347/3881271, then descending to the road junction in the valley on roughly the same alignment as the modern dirt road. No physical remains of the ancient route were observed in the landscape.
44One of the starting points of this investigation was the observation that Late Roman basilicas are often located on roads.65 The presence of a south-north road passing the monastery of Nikolaos and the basilica of Varnavas and Ilarionas fits into this pattern. To test the alternative hypothesis of an east-west route passing through Peristerona, the present-day road from the crossing of the Serrakhis river at Meniko to Peristerona was surveyed in the field. No traces of an ancient road were found, and neither the lateral alignment nor the longitudinal profile of the road exhibit any of the characteristics associated with Roman road engineering practice.
- 66 Lysandrou, Agapiou 2016, p. 873.
- 67 Ibid.; cf. Bekker-Nielsen 2004, p. 223.
- 68 Kyriakou 2019, p. 48.
- 69 Lysandrou, Agapiou 2016, fig. 2.
- 70 Contrary to the opinion of Deligiannakis 2022, p. 99, who considers the connection close.
45According to Lysandrou and Agapiou, Hellenistic and Roman necropoleis are almost exclusively to be found in proximity to the Roman highway network, at a maximum walking distance of one hour.66 Lysandrou and Agapiou suggest that the few examples of necropoleis outside the above said buffer zones were associated with other, still unidentified, roads.67 This hypothesis is supported by finds of early Christian basilicas seemingly unconnected to the road system.68 The Roman necropolis recorded in Peristerona may be located in proximity to the road running along the Peristerona valley.69 Where ancient roads have remained in use into the early modern period, present-day cemeteries are also often located close to the road-line. The recently discovered luxurious, 4th century, villa north of Akaki is located more than 7 km north of the trans-island road and also seems more likely to be associated with a secondary road.70
- 71 Bekker-Nielsen 2004, pp. 208‑209.
- 72 Winther-Jacobsen 2013, p. 331; cf. Winther-Jacobsen 2022, p. 369.
46The existence of a south-north trans-island road from Kourion across the Troodos is attested by the Geography of Ptolemy (2nd century AD), which gives the distance from Kourion to Lapethos, and by remains of the ancient road on the southern slopes which were still visible in 2002.71 Given the nature of the terrain north of the watershed, the route must have followed the valley of the Karkotis (Karyotis) river, while its exact course has been a matter of conjecture. But as Kristina Winther-Jacobsen pointed out: “Bekker-Nielsen (2004, 210–11), due to the Medieval churches and a khan at Kalliana, places the Roman road on the western side of the river. North of modern Evrykhou, however, traffic up and down the valley is more likely to have run on the eastern side of the river, where there are Hellenistic–Roman settlements and tombs”.72
47On the right bank of the river, some distance below a group of Hellenistic-Roman tombs on a westward-facing slope, remains of a north-south road-line are preserved in a series of farm roads extending as far southward as Eurykhou village, where the road forms the western boundary of the village cemetery (fig. 9). The route is presumably anterior to the tombs and would originally have followed a course further up the slope. A ledge or terrace is visible below the tombs and above the modern highway. The layout and longitudinal profile of the field road below the modern highway are consistent with a Roman date and support Winther-Jacobsen’s argument for placing the Roman road on the right bank of the Karkotis (Karyotis). The gravelled road now visible in the landscape (fig. 10) is a likely candidate for the Roman south-north route connecting Kourion to Soloi and Lapethos (fig. 1), keeping to the right bank of the river and continuing down the valley towards Panagia Skouriotissa. In the upper part of the valley, the line of the road is difficult to identify, as sections of the presumed roadline are overlaid by modern road or housing construction. Further field studies in this region would be helpful.
Figure 9 — Ancient and modern roads in the lower Karkotis valley. Contours at 40 m intervals.
T. Bekker Nielsen © LAF.
Figure 10 — This dirt road northwest of Eurykhou village may preserve the alignment of the trans‑island road from Kourion to Lapethos. UTM 36S 36S 490071/3878978
T. Bekker Nielsen © LAF.
48It is a truism that Roman highways, like their Hellenistic predecessors, were primarily laid out to serve military and political purposes: control, communication and colonization. However, the secure identification of the crossing of the Roman highway in the Peristerona valley – not in the current village of Peristerona with the only attested Byzantine church, but in the area where a river crossing first appeared in the Middle Chalcolithic period – is evidence that the course of the road alongside the central northern Troodos foothills is not a product of settlement patterns of the Hellenistic and Roman period, but is in fact much older.
- 73 Bekker-Nielsen 1993, p. 189.
- 74 Bekker-Nielsen 2004, no. 145, pp. 252‑254; Mitford 1950, p. 87.
49This does not mean that Cyprus is an exception to the general rule. The dendritic nature of the Hellenistic road systems, branching out from the poleis which had formerly been independent kingdoms, clearly reveal that their primary purpose was to provide links between the centre (the polis) and the periphery.73 Only after the Roman conquest were these road systems integrated into a single island-wide network, which included a peripheral coastal highway and an inland trans-island highway (both of which are depicted on the Tabula Peutingeriana) as well as a highway into the Karpas peninsula (which is not shown on the Tabula but epigraphically attested by a milestone found near Agios Theodoros [Çayırova]).74
- 75 Bekker-Nielsen 2004, p. 211.
- 76 Bekker-Nielsen 2004, p. 179.
- 77 Bekker-Nielsen, in press.
50In the quadripartite division of the island described by Ptolemy, the copper smelting in the central Peristerona valley would have been located within the district of Lapethos, as would the Skouriotissa area in the lower Karkotis (Karyotis) valley. Lapethos was connected to the Karkotis valley by a road via the Panagra gap (ca 10 km west of ancient Lapethos) and the modern village of Kalokhorio/Kapouti.75 It would come as no surprise to find a similar route branching off at Kalokhorio and heading southeast towards the Peristerona valley. Coming from the north, the valley offers a natural north-south corridor through the Mesaoria and the foothills, leading directly to the copper smelting operation at Agia Marina Xyliatou-Mavrovouni. Possibly the section between the Orounta gap and Tamassos, which in Roman times constituted part of the trans-island highway, originated as a local road to the copper deposits west of Tamassos. This hypothesis is supported by the observation that the longitudinal profiles of the trans-island highway east and west of the Peristerona valley are markedly different.76 While the eastern section exhibits an uneven profile with a great deal of lost ascent, such as might be expected of a “vernacular” road, the western section shows the even gradient and careful attention to lost ascent characteristic of Roman road surveying practice.77
51On this interpretation, a dendritic system of routes – going back to the early decades of Ptolemaic domination, possibly further – connected Lapethos to the mining districts of Skouriotissa and Tamassos, including Agia Marina Xyliatou-Mavrovouni. After the Roman conquest, these routes were integrated into the overall highway network and, where possible, realigned and resurfaced to Roman standards, while the road from Tamassos was extended from the Peristerona valley to Soloi, establishing a trans-island link from the bay of Salamis to the bay of Morphou and facilitating copper exports through the important port at Soloi.