• 1 The literature on Chaleion is collected by Freitag 2000, p. 107‑110.
• 2 For the forms of the name as recorded in the sources, cf. Lerat 1946, p. 331.
• 3 Lawrence 1979, p. 438, n. 9.

1Galaxidi1 occupies the site of ancient Chaleion2, one of the cities in Western or Hesperia Locris that extended from the land of Apollo to Aetolia (fig. 1). This city, that was probably founded during the early years of the Aetolian occupation, in the late 4th c. BCE, survived through the upheaval of the late Hellenistic period and probably also outlived the emergence of Christianity. Its abandonment can be placed with some degree of certainty to the 10th c. CE, when the raiding Bulgarians occupied the settlement and reduced it to ashes. This destruction saw the end of a long history of settlement at the site and resulted in the elimination of almost all the remains of earlier structures, save the fortification wall. The latter, in spite of the hardships it has gone through, is still preserved in a fair condition since almost all of its features are discernible, whereas it functions as the sole landmark of the site. This paper aims to collect the available material on the fortification wall of Chaleion, which is rarely mentioned in handbooks on ancient fortifications3.

• 4 Lerat 1952, I, p. 198‑209.
• 5 According to Galaxidi’s Chronicle, compiled by the monk Euthymios, the most informative text on the (...)
• 6 Spon 1678, p. 46; Leake 1835, p. 593. The travellers are collected by Portelanos 1998, p. 837‑838.
• 7 Dodwell’s text (Dodwell 1919, p. 130‑131) is still useful: The only remains consist of some foundat (...)
• 8 “συνέκειτο ἀπὸ τρεῖς περιβόλους ἐκ λίθων τετραγωνικῶν, ὡς ἐπὶ τὸ πολὺ λελαξευµένων καὶ κανονικῶς ἐπ (...)
• 9 Lerat 1952, I, p. 198‑199.
• 10 Sathas 1865, p. 179.
• 11 Threpsiades 1972, p. 187, fig. 1.
• 12 Papachatzes 1998, p. 459, fig. 474.
• 13 Portelanos 1998, p. 837‑847. For the archaeological research at Galaxidi, cf. Skiadas 1999, 63‑64, (...)
• 14 Skorda 2003, p. 7‑8.

2The identification of the city’s ruins is credited to Lucien Lerat, who, after conducting an extensive survey in Western Locris during the 1940s, discarded the previous suggestions that Galaxidi should be identified with Oeantheia, and suggested instead the correlation between the evidence on Chaleion and the ruins of the modern marine town4. Lerat, through his work published in 1952, not only revised earlier assumptions but also set the foundations for the study of the site’s defense system. The fortification wall that was destroyed long before Lerat visited the region, probably by the late 10th c. CE5, did not attract travelers’ attention, who were content with vague references6. Among them, the most informative is Dodwell’s description7 and that of Sathas, the editor of Galaxidi’s Chronicle, who explicitly states that the wall consisted of three fences made of rectangular stones8. Both the aforementioned testimonies were reported by Lerat9. There is only one explicit testimony as to its destruction: according to Sathas, the wall was demolished in 1830 to provide material for the queue‑side of the modern harbour10. Lerat, with no information from pre‑existing surveys, traced out a new course for the wall on an aerial photo, over which he drew the outline of the plots and the course of the defensive construction as far as possible when not covered by later buildings (fig. 2). Not much was added to Lerat’s work by Travlos’ plan that was published by Threpsiades11 and reproduced by Papachatzes in his translation and commendation of Pausanias’ travels12 (fig. 3). In 1998, Portelanos, in his unpublished dissertation, collected the available material and described the wall, relying mainly on Lerat’s work13 (fig. 4). A final contribution was offered by Skorda (fig. 5), who in 2003 published a plan of the site14. Relying on all these earlier studies, we collected all the documentation yielded during the works of the Archaeological Service and here, an updated version of the Chaleion’s wall is presented (fig. 6).

• 15 Threpsiades 1972. The clay lamps and the glass vessels were studied and presented by Zaphiropoulou (...)
• 16 According to Skorda 2003, p. 8, most important was a jar‑burial that contained thirteen vases and e (...)
• 17 Petrakos 1972, p. 375, pl. 317b; Petrochilos 2017, p. 41, pl. 2 (the mirror). The mirror, in spite (...)
• 18 The first chamber is located in the grove south of the harbour; in the interior, that measures 4.15 (...)

3Regarding archaeological research on the site, the results are not much more extensive. In 1940, Threpsiades excavated a rectangular funerary monument to the west of the settlement, as well as a number of cist and tile graves dating to the Hellenistic and Roman periods (fig. 7). Unfortunately, due to the loss of the excavation journals, the context of each grave is not known15. In 1990 this cemetery was further excavated and six more tile graves were found16. In addition to the western cemetery, a second burial ground, probably related to one of the roads leading to one of the gates, is located to the south of the settlement. One grave was found in 1972 at a distance of approximately 200 m from the enclosure, but the exact circumstances were not recorded. The grave was lavishly furnished with a bronze folding mirror bearing the head of Aphrodite in relief (Mus. Gal. no. 345), a bronze kantharos (Mus. Gal. no. 346), a handle (Mus. Gal. no. 347), a needle (Mus. Gal. no. 348), an undecorated amphora (Mus. Gal. no. 349), a lid of a pyxis (Mus. Gal. no. 350), a black-glazed lamp (Mus. Gal. no. 351), and an unguentarium (Mus. Gal. no. 352)17. Other than a small number of engraved and figure tombstones, two more natural caverns, which were converted into burial chambers with sarcophagi, observed and described by early scholars, can be added to these funerary monuments18.

• 19 Lerat 1952, II, p. 29‑32; IG IX.I2 3, 718. Domínguez 2013, p. 457‑461; Daverio 2015, p. 186‑190.
• 20 Threpsiades 1972, p. 201‑205; Morgan 1990, 254‑256.
• 21 Petrochilos, forthcoming, p. 44.
• 22 Sathas (1865, p. 109) records the narrations of older men who recalled that the humble antiquities (...)
• 23 The inscription on the treaty between the two Locrian cities bears an explicit reference to a harbo (...)

5The city’s history predates its foundation. An epigraphic testimony dated to the early 5th c. BCE, which is acknowledged as one of the most informative legal texts on the establishment of a city with the initiative of a federated state, bears reference to a group of Chaleians. The latter, under the leadership of a certain Antiphatas, along with other people from Hypoknemidian Locris, were sent to settle Naupactus19. This early 5th c. BCE testimony serves as a strong argument in favour of the existence of Chaleion as a political community long before the establishment of the city in modern Galaxidi. Where this community was settled is a matter of speculation. Nevertheless, there are strong indications in favour of the assumption that the Chaleians occupied the slopes in the mountainous area of Galaxidi, where several archaic settlements have been located20 at Palaiogalaxido, where a precipitous, fortified acropolis offered safe conditions for the population living in the surrounding plains, and mainly at Ayios Vlassios21. The latter is beyond a place where an archaic settlement22 developed. Two inscribed plaques, currently on display at the British Museum, are reported to have been discovered; one records a treaty between Chaleion and Oeantheia, and the second is the aforementioned inscription with the regulations concerning the colonisation of Naupactus. These epigraphic findings, as well as several bronze vessels that are reported to have been found at the same site and are stored in several museums abroad, point to the religious character of the site23.

• 24 A suggestion put forward already by Sathas 1865, p. 109. For the phenomenon of the “metoikesis”, cf (...)
• 25 Skorda 2003.

6The relocation of the local population24 and the foundation of the city coincide with the occupation of the region by the Aetolians, who must have had an active involvement in these events. The Aetolians – or in any case the authority that conceived the foundation of the urban planning – understood the geographical advantages of the site. The city is located on a low piece of land jutting into the sea; the settlement could be easily protected from the hinterland, and the two natural harbours, the deeper Limen (Λιµήν) and the shallower Cherolakas (Χηρόλακας), offered ideal conditions for the safe mooring of boats (fig. 8). This latter feature must have been the most decisive for the Chaleians in choosing their city. A partly preserved wall found underwater in the natural harbour of Cherolakas was probably the foundation of a quay stretching from the land into the sea25.

• 26 Grainger 1999, p. 94‑95.
• 27 Grainger 1999, p. 94.
• 28 Grainger 1999, p. 96.
• 29 Grainger 1999, p. 95.

7The Aetolians had already been gradually expanding since the early 5th c. BCE. Initially, they occupied Molykreion, Makyneia, and the suburbs of Naupactus26, the western part of the region hence named as Αἰτωλία ἐπίκτητος. In 338 BCE, Naupactus was also handed over, for a short period, to the Aetolians by Philip27. Only a few years later, in 321 BCE, the Aetolians invaded Western Locris and plundered the territory of Amphissa, a city to the east of Chaleion. Soon they withdrew, forced away by the Acarnanians28, but their policy to occupy Western Locris permanently would simply be delayed for a few years. Prior to the struggle against the Celts in 279 BCE, the Aetolians had established themselves over the entire former Hesperia Locris29.

• 30 Vatin 1968, p. 35, dated to the end of the 4th c. BCE (according to Lerat) or the beginning of the  (...)
• 31 Valavanis 2015, esp. p. 119‑120.

8The foundation of Chaleion at the site of modern Galaxidi seems to have sparked an interest in protecting the hinterland. The density of sites in the region and especially the number of towers indicates that Chaleion was not negligible for the territory’s sovereign. At that time, borderline disputes with Tritaea were settled, as indicated by a very detailed and informative decree on the agricultural occupation of the inhabitants30. In brief, in the wider region, eight towers have been traced that could be approximately contemporary to Chaleion. Opposite Chaleion in the Gulf of Kirra, on a rocky island, there is a wide trench bisecting the terrain which has been recognized as a ship shed, probably belonging to the Aetolians31.

• 32 Lerat 1952, I, p. 138‑146, fig. 6, pl. XXXIII-XXXIX; Portelanos 1998, p. 725‑733, pl. 144‑147, plan (...)
• 33 Lerat 1952, I, p. 99‑101, fig. 2, pl. IX‑X; Portelanos 1998, p. 629‑637, pl. 119‑120, plans 161‑166
• 34 Lerat 1952, I, p. 145‑149, fig. 7, pl. XLI-XLIII; Portelanos 1998, p. 857‑861, pl. 185‑186, plan 23 (...)
• 35 Lerat 1952, I, p. 169‑173, fig. 9, pl. LIV-LVIII; Portelanos 1998, p. 851‑856, pl. 183‑184, plan 23 (...)

9Chaleion’s fortification wall, which is built of tough local limestone, lies on a limestone promontory with gradual slopes rising up to 50 m above sea level. Clear traces of the circuit’s course have been traced on all the sides and located in several parts of the modern city. Following the contours of the hillock and enclosing an area measuring roughly 70,000 m², the wall that was extended to approximately 1000 m was studded with rectangular towers on all sides. The best-preserved parts are those on the NW and NE sides, where the curtain and its towers are visible to a height of approximately 60 and 70 m respectively. The wall’s width measures 2.9 m, with only slight variations, and it is double-faced with quadrilateral rectangular (ashlar) blocks on the faces. In order to adapt to the cavities of the bedrock, the first two courses vary in height, but then they are of equal height (fig. 9), forming a neat, trapezoidal isodomic or pseudo-isodomic masonry that resembles the masonry at Ayioi Pantes (Vidavi)32, Eupalion33, Pendeoria34 and Myonia (Ayia Euthumia)35, all settlements that are located in the region of Western Locris. The fill between the two revetments consists of earth and chipped or unworked piles of stone, into which headers are occasionally inserted (fig. 10). The blocks have quarry faces; their exterior faces received no substantial working besides the one carried out during their extraction at the quarry, or hammer faces, since they are also roughened by means of a blunt implement. Regarding the joints, most are drafted but not bevelled.

• 36 Adam 1982, p. 48, fig. 17; Portelanos 1998, p. 175‑192, pl. 16‑21, plans 25‑28.

10Ten towers have been traced so far, either entirely or partially documented. Their placement does not follow a stable rule and the length of the curtain between them varies considerably. As expected, the towers do not reinforce the points where the stretches of the wall join inwards, thus shaping a kind of recession. None of the extant towers have been excavated so only preliminary remarks can be made on their size. The most modest is tower a on the eastern side, whereas the most extensive is tower h. It seems that their bases were solid and floored with compressed earth and rough stones, as was the case of the towers at Plevron36. There are some indications that partition walls were used to enhance the construction’s massiveness, as well as to prevent the outward thrust of the fill, especially in tower a. Only one gate has been found so far on the south side of the wall. It is an axial gate with one tower flanking to the left, lacking any courtyard.

• 37 Lerat 1952, I, p. 153.
• 38 Cf. Lerat 1952, I, p. 153‑154, fig. 8, no. 2, pl. XLV, 4 and XLVI, nos 1 and 3; Threpsiades 1972, p (...)
• 39 Threpsiades 1972, p. 186.
• 40 The plot was excavated in 1976; 32 m of curtain preserved up to 2 m high (two courses) and 3 m wide (...)
• 41 Skorda 1998, plan 1, pl. 153γ.

11Lerat recognized a possible trace of a tower on the eastern side of the circuit, of which nothing is visible nowadays37. The most conspicuous part of the curtain is approximately 60 m long and 2.90 m wide, with two rectangular towers; tower a is 7.4 m long and protrudes 4.20 m from the curtain; tower b that was excavated in 2018 is of the same dimensions as the latter (fig. 11). The curtain is preserved up to four courses (two metres) in height. One of the plinths bears a relief in a frame resembling a shrine that contains two arched symbols; Lerat interpreted them as Dioscuri38. Between parts [1] and [2] a considerable portion of the circuit is missing, approximately 75 m long. Travlos acknowledged traces of the wall towards the tip of the promontory39. On the other hand, along the circuit’s N course Lerat found some remnants at three spots (his number 3), which are not visible today. These indications suggest that following the tower b, the wall turned abruptly to the NE towards tower c, at a distance of 110 m. Travlos was probably carried away by some rock-cuttings and ruins of walls preserved towards the edge of the promontory, next to the modern pier. These were also noticed by Lerat, who attributed them to a detached bastion. Since there are no more traces of the structure to confirm this suggestion, one need not discard it. Of the expected 110 m long N side, where one would expect a tower, 13 m of the curtain wall have been revealed and are still visible at the foundation level of a modern building. This part of the enclosure was reinforced by tower c that measures 5 m in length and projects from the curtain 2.80 m40. The retreat of the precinct’s outline from the perimeter of the promontory, contrary to the reconstruction put forward by Travlos, would leave a flat, quite extensive area in front of the N side, empty, an observation that needs further discussion. This part of the promontory is, on account of its tight contours, the least accessible from the sea, so it would be rather difficult to attack this side. On the other hand, a courtyard in front of the wall could facilitate an unhindered approach to a gate. A small staircase carved on the SE side could lead to an open terrace, in close proximity to tower b. Indeed, it seems that in the middle of the north side, there had been an opening, part of which was revealed during the excavation at Psychogiou’s plot (fig. 12). Nowadays only the western part of the threshold, as well as the pilaster of the same side, is visible. It is worth mentioning that in the contemporary town planning, there is a path whose course coincides with the ancient opening at the circuit. The NW corner of the fortification perimeter was buttered by the rectangular tower d, which has not been fully excavated. There then follow 29 m of curtain measuring 2.90 m in width [3] and after an interval of 18 m there is another tower, e, measuring 6.10 m in length and 4.20 m in width. From its NW side almost 40 m of curtain extend [4] that after an interval bents abruptly to the east towards the harbor [5]41. Then the curtain follows a course almost parallel to the shore for approximately 55 m, of which 23 m are preserved to foundation level [6]. Upon the latter there were subsequent additions built of quadrilateral rectangular blocks and plinths. At the middle of the north side of the circuit the exterior face of the wall was founded in a groove that was carved into the rock for more than 9 m, whereas the interior fill was based on the rock that was almost untouched, other than some slight configurations.

• 42 Skorda 2003.

12Thereafter the circuit follows a southern almost straight course for more than 100 m that must have been strengthened with at least two towers. From one of them, tower f, the remains, merged into the cavities of a rock, are preserved (fig. 13). This part should not be identified as a buttress since the preserved stretches of the circuit show no evidence for others. Only the northern face of the tower is preserved, whereas the other sides are dismantled. Almost in the middle of the western stretch, the fortification wall is penetrated by a gate 2.30 m wide [8] (fig. 14). At both ends, there are pivot sockets for the door leaves, measuring 0.20 m × 0.10 m. The threshold consisted of three slabs, of which the central one is missing. An alternative explanation is that the middle slab was never placed, to help drain water from the city. north of the gate, on the left‑hand side for the person approaching, tower g measures 7.20 m long and 3.25 m wide. From the western face of the circuit two more parts [9‑10] are preserved. At this point, Travlos assumed that the wall’s course ran further down the slope to the west of the extant remains, where he also reconstructed a recession penetrated by a gate (his number 5). Nevertheless, there are no indications of the fortification at this location. Tower h, being 12.40 m long but rather narrow (4.70 m on the southern and 2.50 m on the northern face, an asymmetry owed to the oblique direction of the south-western course of the wall) was the most extensive tower. On the southern face, three parts of the wall are preserved along with tower i, which is 7 m long and 4.20 m wide. The south-eastern corner of the circuit is reinforced by tower j42. Two stretches of the wall are preserved, one between towers i and j [17], and a second [18] to the north of the latter.

• 43 This method of dating the fortifications was formulated mainly by Wrede 1933 and Scranton 1941. The (...)
• 44 A notable exception is Lawrence 1979, p. 438, n. 4, who justifies his assumption that the absence o (...)
• 45 Pappadakis 1923, p. 147. Lerat, Chamoux 1947/8, p. 56‑57; Papachatzes 1998, p. 458‑459.
• 46 Themelis 1978, p. 147 and 1984, p. 232.
• 47 Portelanos 1998, p. 846; Skorda 2003, p. 8; Valavanis 2015, p. 120.

14Dating a fortification wall is an arduous task, especially when the excavation finds related to it are not recovered from undisturbed contexts. The lack of safe documentation is not compensated by morphological evidence or the type of masonry43. As for the date of the Chaleion’s wall, the majority44 of scholars who have studied it postulate a late 4th or early 3rd c. BCE date. Lerat, following Pappadakis45, assumed that it was erected after the Aetolian occupation of the region. In 1978, Themelis, while removing the fill from the cavities in front of the wall in the southeastern part of the wall (former Chatzis’ plot), recovered fragmented pottery dated to the late 4th or early 3rd c. BCE (320‑290 BCE)46. This dating has been accepted in the few ensuing references47.

• 48 Garlan 1974, p. 19‑86; McNicoll 1986.
• 49 Garlan 1974, p. 82.

15Generally speaking, the size and the features of the Chaleion’s wall indicate that its date in the 3rd c. BCE needs to be excluded since this does not take into account the developments that had already taken place in the 4th c. BCE. At that time, the introduction of the mechanical siege-craft techniques caused the re‑appraisal of the hitherto defensive strategy48. Until that time, the principles set in the archaic period were still valid and the fortification walls continued to employ the concept of the simple, vertical barrier that dated back to the prehistoric period and frequently protected only the acropolis of the settlement. Probably in the 6th BCE, the addition of the tower as a means to enhance the defenders’ ability to counterattack was introduced. In the 5th c. BCE, the defensive wall surrounded the entire settlement and later in the 4th c. BCE even more extended constructions were built to enclose the non-inhabited areas or the pasture land pertinent to the city’s sustenance. This so‑called Geländemauer type of circuit was considered in the Classical period an adequate defense against the contemporary siege-craft method49.

• 50 Winter 1971, 123; Lawrence 1979, p. 39.
• 51 Marsden 1969, p. 48‑56; Winter 1963; Ober 1987; Garlan 1974, p. 156‑168.
• 52 Ober 1987, p. 579.
• 53 Lawrence 1979, p. 223.

16In addition to that, the principles on which the defense systems of the Classical period were structured were reviewed, when in the early 4th c. BCE Dionysios of Syracuse employed effective catapults that eventually brought about revolutionary changes in war tactics and the concept of defense. Walls that were built until Dionysios’ time were thenceforth deemed obsolete (ἀρχαία is the term applied by Philo when he refers to the defense systems prior to the Hellenistic period, which consisted of continuous walls that were interrupted only by projecting towers)50 since they could not accommodate the mechanical implements his engineers had invented. At Syracuse, he also assigned his architects to construct a larger wall that would be defended by higher towers with windows on the upper level, from where the warriors could light catapults against the assailing foes51. The catapult’s power was directly proportional to the size of their bows and eventually to the available space on the tower52. These new standards were soon to be followed in the newly founded cities, a demand put forward already by Aristoteles in his Politics (1330 B‑1331 A), who advocated in favour of erecting more secure walls in view of the advances made in the field of military artillery; from the transition to the 3rd c. BCE the towers became sturdier, larger, and heavier53.

• 54 Winter 1989, p. 191.
• 55 Winter 1989, p. 156, 165‑7, 323‑4 on defensive artillery.
• 56 Ober 1987; but also the reservations and amendments of Rihll 2006.
• 57 Winter 1989, p. 192, n. 8.
• 58 Krause 1972, p. 72; Adam 1982, p. 206, fig. 120.
• 59 Adam 1982, p. 216, fig. 126, and Idem 1992, p. 14, fig. 1.6.
• 60 Themelis 1984, p. 237; Pantos 1985, p. 462‑463, based on historical grounds and on the account that (...)
• 61 Lerat 1952, I, p. 189‑191, fig. 1, pl. I, II; Portelanos 1998, p. 304‑319, pl. 46‑48, plan 80.
• 62 Portelanos 1998, n. 33.
• 63 Portelanos 1998, n. 34.
• 64 Portelanos 1998, n. 32.
• 65 The fortification structures running around small settlements (kômai) in Portelanos 1998, p. 1457‑1 (...)

17The wall at Chaleion does not conform to the developments recorded so far regarding military architecture, and in this sense, it seems to have been structured on rather obsolete principles. The circuit surrounds a rather restricted area and does not enclose non-inhabited arable areas or pasture. Besides that, the towers’ size does not correspond to an advanced military strategy, in which the machines employed tend to become larger. The first generation of circuits, after the invention of torsion artillery and related equipment, was small scale and the chambers inside the towers where correspondingly of modest scale. The torsion artillery would not have demanded a space larger than 3.5 m deep beyond the curtain wall, other than the wall itself, making an area of 25 m² in total54. Chambers of that size could have accommodated 1 ½ – mina stone-throwers55 that required a clear space of 3.3 to 3.4 m both for the machine and the operator. Later on, as bigger and more powerful artillery was built, the former towers would have been considered as inadequate56. Towers large enough to accommodate torsion stone-throwers are rare before the end of the 4th c. BCE. The aforementioned remarks are not necessarily conducive to the dating of the wall, since smaller towers were employed even in the Hellenistic period57. In any case, the size of the towers alone is not an adequate indication for the date of a fortification. Nevertheless, the small size of the towers suggests an earlier date. Furthermore, the gate without courtyard or overlapping curtains that would force the enemy to crowd into a very limited space while approaching, is the simplest form of gate that illustrates the most rudimental principles of defense found in many places from the prehistoric (as in the case of Troy’s VI phase) to the late Classical period, as in the case of Phyle’s fort58 or that at Eleutherai (Gyphtokastro)59. In general, similar features, such as those of Chaleion, can be seen at the fortification circuit at Kallipolis or Kallion, which was probably erected towards the end of the 4th c. BCE60. Finally, probably due to the terrain formation, where there were no inaccessible places large enough to shelter the population in times of danger, the founders were obliged to obviate the construction of an acropolis. This feature is observed in contemporary settlements in the region, as is the case of Makyneia61, Eupalion62, Pendeoria63, Ayioi Pantes (Vidavi)64, to name just a few65.

• 66 Galaxidi Museum no. 544. SNG Cop. 533.
• 67 Portelanos 1998, n. 17.

18Apart from the aforementioned general observations that rely on the military advances in the Classical and Hellenistic periods, the moveable finds from the site corroborate the date proposed by earlier scholars. It is questionable if a stray find that was delivered to the Museum of Galaxidi, a bronze coin of Aegina bearing two dolphins on the obverse and the refined skew pattern in the incuse square on the reverse, dated to the second quarter of the 4th c. BCE66, was actually found at the site. As already noted, the earliest tomb found at Galaxidi and published by Petrakos in 197267, is dated to the second half of the 4th c. BCE.

• 68 Rotroff 1997, p. 247 no. 52, fig. 7, pl. 6 (ca 325 BCE).
• 69 Rotroff 1997, p. 309‑310, no. 641, fig. 46, pl. 60 (325‑300 BCE).
• 70 Rotroff 1982, p. 88, no. 373, pl. 66, 87 (middle of the 2nd c. BCE).
• 71 Rotroff 1997, p. 362, no. 234, fig. 75, pl. 90 (200‑175 BCE).
• 72 Rotroff 1997, p. 334, no. 918, fig. 60, pl. 73 (225‑175 BCE).
• 73 Rotroff 1997, p. 258, no. 154, fig. 12, pl. 14 (300‑275 BCE).

19Additionally, the re‑examination of the pottery and the other small finds recovered from the foundation level of the fortification enclosure helps to narrow down the date. Thus, the tapering stem from a kantharos that was found in the foundation trench [no. 2 on the plan] is probably dated to ca. 325 BCE68 (fig. 16), whereas the partly preserved black-glazed plate with palmette impressions within rouletting from the same place is also dated to the last quarter of the 4th c. BCE69 (fig. 17). Much more abundant is the middle and late Hellenistic material, as well as that of the Roman imperial period; e.g. the fragment of a relief bowl70 (fig. 18), the base of a pyxis71 (fig. 19), the base of a small bowl72 (fig. 20), the small attic‑type glazed skyphos73 (fig. 21). Taking into account the archaeological documentation, the Chaleion’s wall was probably founded towards the earlier years of the last quarter of the 4th c. BCE.

• 74 IG IX,1² 3:724; 725; 726; 729 (3rd c. BCE); 727; 728; 732; 733 (2nd c. BCE); 730; 731; 734; 735, 73 (...)
• 75 IG IX,1² 3, 723.
• 76 Undecorated jug, inv. no. 95 (from Kavos).
• 77 Unpublished. Plain slab without border. Missing the left bottom corner with the lower left part of (...)

21As already mentioned, the uninterrupted habitation of the site resulted in the almost complete elimination of the earlier constructions. This observation accounts for the preservation of only a handful of inscribed gravestones and even less sculptural works. None of them predates the early 3rd c. BCE and most are dated to the 3rd and 2nd c. BCE74. The earliest funerary inscription is the one for Harmoxenos, whose stele was erected by his father and mother as a memorial of their child to their descendants, as we learn from the epigram inscribed on the base of the monument75. The shape of the letters, especially the alphas with the broken vertical bars and the sigmas with slightly divergent horizontal bars, finds parallels to inscriptions of the early 3rd c. BCE from Delphi. Besides Threpsiades’ and Skorda’s excavations in 1940 and 1990 respectively, which brought to light Hellenistic finds, the random findings of that period are scarce. Objects dated to the Roman period are more frequent. Apart from the Roman tile graves excavated by Threpsiades in 1940, vessels from Galaxidi have been delivered to the Archaeological Service since before the middle of the 20th c.76 Only two sculptural works have been found so far at Galaxidi, the funerary stele of a girl, probably dated at the late 2nd c. CE77 (fig. 23) and the lower leg of a life‑sized statue of a man, also dated to the Roman period.