• 1 Pikoulas 1990‑91, p. 247‑248; Pikoulas 1995, p. 28.
• 2 A.W. Lawrence, Greek Aims in Fortification (1979), p. 188.

3The term watch-tower has been chosen for the structures under discussion as they are isolated buildings outside the urban plan of a city‑state’s territory. According to ancient literary sources for similar buildings different names have been assigned to these types of structures, such as πύργοι, φρυκτωρίες, φρυκτώρια, πυρσουρίδες, πυρσούρια, φυλάκια, φυλακεία, φυλακτήρια, σκοπές, φρούρια1, and περιπολεία2. Furthermore, according to Pseudo-Demosthenes (47-56) a tower may be characterized as the fortified part of a farmhouse.

• 3 F. Dakoronia, AD 50 (1995), Chron., p. 341‑342, pl. 4.

4A fire in a public forest area near the modern village of Sfaka in the Municipality of Amfikleia-Elateia, in the summer of 1994, revealed the remains of a watch-tower (fig. 2). A rescue excavation was conducted and verified the character of the building3.

• 4 A. Orlandos, Τα υλικά δοµής των αρχαίων Ελλήνων (1959‑1960), p. 220‑222.
• 5 F. E. Winter, Greekfortifications (1971), p. 71‑77; Pikoulas 1995, p. 330.

6The watch-tower excavated in 1994 is located 19 km from Atalante, on the modern road from Atalante to Elateia. It is square in plan measuring 9 × 9 m and its masonry is pseudoisodomic trapezoidal (fig. 3)4. Local large square‑cut limestone blocks were used for its construction, laid on the hard rocky ground, prepared for this purpose (fig. 4). The walls are preserved to a maximum height of 0.92 m. Their upper surface was shaped to receive further courses on top, a fact also deduced from the scattered ashlar blocks discovered throughout the building. There is no evidence of the final height of the building or of how many upper floors there may have been. The use of other building material for the wall superstructure, such as mud‑bricks and timber cannot be excluded, as suggested by other comparable examples5.

• 6 S. Morris, « The Towers of Ancient Leukas. Results of a Topographic Survey, 1991‑1992 », Hesperia 7 (...)

7The structure is divided by an inner wall into two compartments. The smaller one to the north measures 8 × 3 m. On its eastern side, another wall, constructed of roughly-worked medium stones, forms a small area, which could be interpreted either as an air‑duct or a storage room6. The position of the main entrance of the building is not clear due to the poor preservation of its ground plan. However, a gap on the outer north wall could be interpreted as such. On the same axis in the middle of the partition wall there is a 1.05 m wide opening, connecting the two compartments. The tower was roofed with tiles of laconian type, as indicated by the destruction layers of the northern part.

• 7 A. Orlandos (n. 4), p. 221.

9The building techniques attested at the watch-tower of Sfaka can be compared with those of numerous contemporary architectural remains recorded in the region under discussion, and dated to the 3rd cent. B.C.7, a period when we believe the watch-tower was erected. This suggestion is reinforced by the pottery collected during the excavation, ranging chronologically from the 3rd cent. B.C. to the beginning of the 1st cent. B.C. when the building was probably destroyed.

• 8 Fig. 05a, see Rotroff 2006, p. 256, no 112, fig. 18, pl. 117, date: 225‑190 b.C. Fig. 05b, see Rotr (...)
• 9 Fig. 5c, seeRotroff 2006, p. 304, no 568, 569, fig. 72, pl. 61, date: 175‑150 b.C. & 150‑110 b.C. r (...)
• 10 Fig. 5d, seeF. Dakoronia, « Σύνολα κεραµικής από τάφους µε νοµίσµατα από την Ανατολική Λοκρίδα », i (...)
• 11 Fig. 5e, see Rotroff 1997, p. 349, no 1108, fig. 68, date: 200‑175 b.C.
  Fig. 5f, see S. Rotroff, He (...)
• 12 Fig. 5g, see G. Zachos, « Ελληνιστικοί τάφοι από την Αρχαία Τιθορέα (Φωκίδα) », in Kazakou 2004, p. (...)
• 13 Pikoulas 1990‑91, p. 252.
• 14 Pikoulas 1995, p. 358.

10Pottery mainly consists of vessels for storage (figs 5a, 5b)8, cooking (fig. 5c)9, eating (fig. 5d)10 and drinking (figs 5e, 5f)11. A few oil‑lamps were also discovered (fig. 5g)12. These are ordinary household utensils necessary for the needs of the garrison. These types of artefacts do not contradict the identification of the building as a watch-tower since there seems to be no typical military pottery13. Sparta is an exception to this rule, where the vessel κώθων is referred as part of a soldier’s equipment (Aristophanes, Hippeis 600)14.

• 15 H. O. Schmitt, Die Angriffswaffen, in R. Felsch (ed.), KALAPODI II. Ergebnisse der Ausgrabungen im (...)

11The identification of the excavated building at Sfaka as a watch‑tower is also supported by the discovery of an iron spearhead (fig. 6) whose type is attested at the nearby shrine of Kalapodi and is dated at 4^th-2nd cent. B.C15.

• 16 F. Dakoronia, P. Kounouklas, « ∆ρόµος µε παρελθόν », in International Conference. Ancient Phokis. N (...)

12Α survey conducted on the surrounding area of the excavation, after a recent fire, enriched the archaeological record with more architectural remains consisting of stretches of wall with similar masonry to that of the excavated watch‑tower, and probably belonging to a similar type of building (fig. 7)16.

• 17 For the first reference of the existence of a small part of an ancient paved road next to the watch (...)
• 18 Pikoulas 1995, p. 17.
• 19 A. Yialouri, « Ερείπια κτιριακού συγκροτήµατος στο Καλαπόδι Φθιώτιδας », in A. Mazarakis (ed.), 3^ο (...)

14The ancient road runs parallel to the modern road of Elateia-Kalapodi-Atalante17. Paved roads are constructed on the Greek mainland at the beginning of the Roman conquest, probably in the middle of the 2nd cent. B.C18. The presence of a modern road following the ancient one confirms that ancient road networks remained in use until modern times and change only when new construction principles and economic parameters make it essential. An inn and baths dated to the early Byzantine period, discovered during public works for the widening of the modern Elateia-Kalapodi-Atalante road, add further evidence19. Such buildings constructed near roads were useful for the travellers.

• 20 A. Orlandos (n. 4), p. 71; F. E. Winter (n.5), p. 77.

15The preserved part of the ancient road measures 18 m in length and its width ranges from 1.5 to 2 m. For its construction local limestone slabs of different sizes were used, similar to the building material of the watch-tower20.

• 21 Zachos 2004, p. 208; ead., Ελάτεια Ελληνιστική και ρωµαϊκή Περίοδος (2013), p. 37.

16The two above mentioned buildings are situated on the top of low hills on both sides of the ancient road, which follows the saddle between the summits of Velaora and Prophitis Ilias, both of which were within Elateia’s eastern-northeastern boundaries21.

• 22 S. Prignitz, « Zur Identifizierung des Heiligtum von Kalapodi », ZPE 189 (2014), p. 141‑143.
• 23 S. Prignitz, « Zur Identifizierung des Heiligtum von Kalapodi », ZPE 189 (2014), p. 144‑146.

17Based on excavation and epigraphic evidence it is possible that this road can be identified with the one referred to by Pausanias (X, 35, 1) leading to the district governed by Abae22, the Phocaean city which housed the Shrine and Oracle of Apollo,23 according to ancient literary sources.

• 24 G. Zachos 2013 (n. 21), p. 119.

18The date of erection of Sfaka’s watch‑tower to the 3rd cent. B.C. is backed up by the fact that this period is characterized by repeated conflicts between the Greek ethne. The pottery collected allows us to date the destruction of the watch‑tower to the first half of the 1st cent. B.C. and more precisely during the First Mithridatic War, some of which took place in the region of Elateia24. The view that Sulla, during his campaign against Greece, destroyed the Greek cities and their fortifications cannot be true in the case of Sfaka’s watch‑tower, since the Elateians remained pro-Roman, and perhaps helped the Roman victory at the battle of Chaeronea. This friendly attitude was rewarded by the Romans with the declaration of Elateia as a free, and not taxed, city (civitasliberaetimmunis), the only city in fact that fought on the side of the Romans.

• 25 Y. Bequignon, La vallée du Spercheios (1937), p. 5‑6; F. Dakoronia, Μάρµαρα. ΤαΥποµυκηναϊκάΝεκροταφ (...)

19In the antiquity the territory of a city‑state included the area of its economic interests and activities, which were usually dictated by the topography. A similar situation is still true for modern Greek villages, where their boundaries extend to the land belonging to the members of the community, whilst the rest is public woodland25.

• 26 Pikoulas 1990‑91, p. 253‑254.

20The two buildings, located at both sides of the ancient road, can be characterized as roadside guard-houses. Such structures are located on the outskirts of the city‑states, on the borders where sovereignty was not normally exercised, at locations far from arable land, or rocky areas suitable only for pasture26.

• 27 K. Tsaousi, “Οδικό δίκτυο και άµυνα στη ∆‑Β∆ Αττική”, Thesis, University of Thessaly (2003), p. 6.

21From the ancient literary sources we know that the city‑states kept garrisons in such watch-towers to prevent invasions or any kind of hostile activity. They also protected travellers from attack, prevented cattle stealing and sometimes functioned as custom-houses27.

• 28 Zachos 2004, p. 212‑214.

22A study of the topography of ancient Elateia confirmed the existence of a network of watch-towers for the protection of the city and its countryside28. The buildings of Sfaka may have belonged to the defence system of Elateia. This is evidenced by the geographical position of the structures, whose visibility is unlimited towards the territorial area of the city. They were constructed at the borders of Elateia’s territory, perhaps, on the basis of a central authority’s plan, rather than as private enterprises.

• 29 A.W. Lawrence (supra, n.2), p. 189; Pikoulas 1990‑91, p. 250.

23The planning of such networks, which constitute the defence system of the countryside, is affected by the geographical relief29. The buildings of Sfaka must therefore have served as frontier guardhouses for the Elateian territory.

• 30 F. Dakoronia, AD 40 (1985), Chron., p. 175‑176.

25An excavation undertaken in 1985 revealed a rectangular building measuring 9.80 × 7.20 m founded on rocky ground (fig. 9)30. Its walls were built with rectangular roughly-worked limestone blocks in pseudoisodomic masonry (fig. 10). They are preserved up to four courses to a maximum height of 2 m.

• 31 Fig. 11a, see H. Robinson, Pottery of the Roman World. Chronology, AthenianAgora V (1959), group F, (...)
• 32 Fig. 11b, see H. Robinson, Pottery of the Roman World. Chronology, AthenianAgora V (1959), group F, (...)
• 33 Fig. 11c, see Rotroff 1997, p. 271, no 276, pl. 28, date: 150‑100 B.C.
• 34 Fig. 11d, see Rotroff 1997, p. 317, no 730, fig. 51, pl. 65, date: ca 175 B.C.
• 35 Figs 11e, 11f, see Rotroff 1997, p. 17, no 61, pl. 74, date: end of 2nd cent. B.C. and p. 83, no 32 (...)
• 36 Fig. 11g, see R. Howland, Greek lamps and their survivals, Athenian Agora IV (1958), p. 120, no 500 (...)
• 37 Bouyia 2000, p. 51‑59.

28Dating the complex to the Hellenistic period and more precisely to the end of the 4th cent. B.C. or the beginning of the 3rd cent. B.C. is confirmed by the pottery. It consists, mainly, of sherds belonging to cooking pots (fig. 11a)31, containers (fig. 11b)32 and drinking (fig. 11c)33 and eating vessels (fig. 11d)34. A few examples of megarian bowls (figs 11e, 11f)35 and oil lamps (fig. 11g)36 have been also recorded. Furthermore the building technique can be compared to that of numerous contemporary fortifications in Eastern Locris37.

• 38 F. Dakoronia, E. Zachou, AD 54 (1999), Chron., p. 370, 371, fig. 25.
• 39 Fig. 14a, see Rotroff 1997, p. 387, no 1484, fig. 87, pl. 112, date: ca 300 B.C.
• 40 Sherds of plain pottery belonging to large closed vessels indicate the presence of this ceramic cat (...)
• 41 Fig. 14b, see Rotroff 1997, p. 415‑416, no 1701, fig. 101, pl. 135, date: Late Hellenistic period.
• 42 Fig. 14c, see Rotroff 1997, p. 324, no 815, fig. 55, pl. 68, date: 110‑86 b.C.
• 43 Fig. 14d, see R. Howland 1958 (n. 8), p. 99, no 431, type 32, pls 15, 41, date: end of 3rd – early (...)
• 44 Fig. 14e, see S. Rotroff 1982 (n.11), p. 45, no 3‑4, pl. 1, date: ca 225‑200 B.C.
• 45 Fig. 14f, see Ch. Lazos, Παίζοντας στο χρόνο. Αρχαιοελληνικά και βυζαντινά παιχνίδια 1700 π.Χ. – 15 (...)

30During illegal forest clearing at the coastal hill of Mikrovivosnear the modern village of Tragana another similar structure was discovered and excavated38. It is rectangular in shape measuring 9 × 8 m with a northwest to southeast orientation (fig. 12). It should be noted that the southwest limit of this building was destroyed due to the above mentioned illegal activities. Its walls were low, preserved to a maximum height of 0.73 m. They were built with roughly-worked stones of medium size, carefully shaped without mortar (fig. 13). Some larger blocks were also used. Inner walls divided the building into smaller rooms. It was roofed by tiles of laconian type, fragments of which were found in the disturbed thin layer of soil that covered the whole structure. It is worth mentioning the similarities in plan, construction and building materials of both the buildings of Mikrovivos and Sfaka. The movable finds of the excavation consist mainly of sherds from cooking pots (fig. 14a)39, containers40, drinking (fig. 14b)41 and eating vessels (fig. 14c)42. A few fragments of oil lamps (fig. 14d)43, megarian bowls (fig. 14e)44 and a rounded sherd were also recorded, the latter probably used as a counter for backgammon (fig. 14f)45. According to the available evidence the use of the building ranges from the second half of the 4th cent. B.C. to the first half of the 1st cent. B.C.

• 46 J. Fossey, TheancienttopographyofOpountianLokris (1990), p. 145, 150, 162; Bouyia 2000, p. 58‑59; F (...)

32The watch-tower of Microvivos is part of a guarding system for the coast of Opountian Locris, interacting with the fortified port of Kynos, and the fortified city of Alope and two other possible watch-towers which have been recently reported46.

• 47 F. A. Cooper, « Epaminondas and Greek Fortifications », ΑJA 90 (1986), p. 195; Bouyia 2000, p. 53.
• 48 Pikoulas 1990‑91, p. 247‑248; Pikoulas 1995, p. 28.

35The buildings of Mikrovivos and Valakatousa (Kastro) seem to have defined the boundaries of Opountian Locris. The period of the erection of the watch-tower at Mikrovivos coincides with the conquest of Boeotia in the southern part of Opountian Locris, as a result of the campaigns of Epaminondas in the 4th cent. B.C47. On the other hand, the Valakatousa (Kastro) building marks the northern limit of Opountian Locris and controlled the natural terrestrial passage of the Dipotamos river towards ancient Phocis. It also had a direct view over the territory of Epiknemidian Locris, a fact which also strengthens the hypothesis for its identification as a watch-tower. Structures overlooking large areas could also function as beacons in order signal imminent danger using fire or mirrors. This is the reason why they are referred to as either beacon-towers (φρυκτωρίες) or watch-towers (φυλακεία)48.

• 49 Pikoulas 1990‑91, p. 249‑250; Pikoulas 1995, p. 360.

36The sites where the aforementioned Locrian structures are located display unlimited visibility to the coastline, the valleys and the passes inland, thus protecting the ships’ routes, transport of vehicles and travellers. Surveys carried out in various regions of Greece (i.e. Peloponnese, Attica, etc.) demonstrate the close relationship between communication networks and defence systems49, which is also confirmed for the area of Eastern Locris.

• 50 Bouyia 2000, p. 56; F. Dakoronia et al. 2002 (n. 46), p. 67.

37The available evidence obtained from the Locrian watch-towers, and in particular their distribution, masonry and the movable finds recorded, suggest that their erection can be ascribed to the defensive programme of Demetrios Poliorcetes for the region of East Locris, a view that has been supported by other scholars50. Their destruction and abandonment, including the coastal settlement of Eastern Locris, is the result of Sulla’s raid (Plutarch, Sulla, 26).

• 51 A. Orlandos (n. 4), p. 71; F. E. Winter (n. 5), p. 77.

38Looking at the question of the building materials used for the construction of these structures, there is evidence of quarrying in the nearby areas, which would have saved time, money and effort51. This is confirmed by all the aforementioned Phocean and Locrian monuments.

• 52 S. Morris (n. 6), p. 338‑343.

39According to the relevant bibliography for these types of isolated structures (watch-towers and beacon-towers), very few have been excavated systematically, although many examples have been attested in the Peloponnese, the Greek mainland and the Islands of the Aegean. Thus, it is not possible to establish a reliable typology of their form and a complete study of their spatial organization. Only in general and for specific areas (i.e. Attica, Argolid and the Islands) has it been possible to suggest some typological proposals for these types of buildings52.

• 53 Pikoulas 1990‑91, p. 252; Pikoulas 1995, p. 329.
• 54 A. W. Lawrence (n. 2), p. 189.
• 55 J.‑P. Adam, L’architecture militaire grecque (1982), p. 72.
• 56 Pikoulas 1995, p. 338

40As for the watch-tower plans, two types can be identified: rectilinear (either rectangular or square) and curvilinear (circular or ellipsoid)53. The circular type is preferred on the Islands, especially at coastal locations54, whilst the rectilinear one prevails in the Peloponnese and Central Greece, without excluding that there are some examples of rectilinear watch-towers on the Islands and some curvilinear ones on the Greek mainland. A sub‑type of the rectilinear watch-tower is that of the pyramidal form, for which we only have two debateable examples in the Argolid55. According to the available archaeological record the shape of a watch-tower cannot be regarded as a criterion for its date56.

• 57 J. McKesson Camp II, « Notes on the Towers and Borders of Classical Boiotia », AJΑ 95, no. 2 (1991) (...)

42McKesson Camp II studied a group of watch-towers located in Boeotia and suggested a typology based, mainly, on their ground plan. He also introduced the term “compartment towers” for the relevant examples with partitions and inner rooms, concluding that this type represents a Boeotian pattern57. The buildings at Sfaka, Profitis Ilias (Megaplatanos) and Mikrovivos can be characterized as “compartment towers”, perhaps influenced by neighbouring Boeotia.

• 58 Pikoulas 1990‑91, p. 255.

44Another typology for these constructions has been suggested based on their location and use. Free standing buildings erected within the boundaries of a city‑state, in areas of strategic importance, on top of hills with unlimited visibility, next to road networks and regions of economic interest, are characterized as watch‑ or beacon‑ towers. Similar structures situated amid plains are considered to be annexes of farmhouses for the protection of the fields, the storage of agricultural production and the accommodation of the farmers. Objections to this view have been expressed on the basis that, unlike the Romans who constructed such individual permanent farmhouses on their farmlands, the ancient Greek, as a highly social creature, was a farmer and worker of the countryside but at the same time lived as a permanent resident and citizen in his settlement. The Greek villagers today still maintain this custom, working all day in their fields and returning to the coffee house (καφενείον)in their village in the evening. Therefore their presence in the countryside was restricted in periods when specific agricultural activities were taking place, such as during reaping or the grape-harvest, and did not require permanent installations58.

• 59 A. W. Lawrence (n. 2), p. 187.

45The watch-towers we are dealing with are free standing buildings that do not belong to farm houses, a type otherwise very rare for mainland Greece59.

• 60 K. Tsaousi (n. 27), p. 7.

47This type of defense system for the countryside of Greece loses its importance and is abandoned after the conquest by the Romans, the announcement of Greece as a Roman Province and the establishment of PaxRomana60. This fact is confirmed by the archaeological data from the three presented watch-towers, all of which cease to be in use during the first half of the 1st cent. B.C.