Late Antique Lamps in Attic Caves: Evidence of Cult Practice?

The Classical period represents the peak of the joint cult of Pan and the Nymphs in Attica. [1] The deities were honored together in the caves located in the asty, on the north slope of the Acropolis and elsewhere, as well as in several caves in the chora. [2] These caves were visited more sporadically during the Hellenistic period, and were apparently abandoned by the Imperial period. Some of them were then reoccupied in Late Antiquity, but scholars disagree about the nature of this reuse. Were the ancient cults revived? [3] Or were the caves visited by Christians? [4] Could the material rather point to “magical” rituals? [5] Whatever their interpretation, all scholars agree that they were used for religious purposes. Yet the various types of archaeological material usually found in ancient sanctuaries is completely lacking, and the Late Antique phase is almost exclusively represented by lamps—in sometimes significant numbers, e.g. up to 2,000 in the Phyle Cave on Mount Parnes. [6] As one of the main lighting devices used in Antiquity, lamps are found in a variety of contexts, including domestic. Therefore, we should not take for granted that the lamps found in Attic caves point to religious rituals, even if they are found in former cult sites.

The purpose of my research at the CHS was the publication of my PhD thesis on the abandonment of sanctuaries in the ancient Greek world, from the Archaic to the Imperial period. As part of this project, I wanted to investigate the presence of Late Antique lamps in caves for two main reasons. First, it is important to identify what type of material constitutes evidence of a continuous or renewed cult practice. If only one type of material is represented for a certain period of time, for instance lamps or coins, is it possible to conclude that the cult is still active? Second, these caves allow for reflection on the memory of abandoned cult sites. Did the population of Attica remember the location of these cult places and the identity of the deities that used to be worshipped there, and does this explain why these caves were chosen as the location of the activities that took place there in Late Antiquity?

Caves where Pan was honored together with the Nymphs in the Classical period that also seem to be gradually abandoned in the Hellenistic or the Early Imperial period include the Cave of the Nympholept at Vari, [7] the Phyle Cave on Mount Parnes, [8] the Cave of Pan at Oinoe, in the Marathonian Tetrapolis, [9] and the Cave of the Nymphs on Mount Pentelikon. [10] The material found in these caves includes numerous terracotta figurines representing women or, less frequently, Pan; reliefs representing Pan and the Nymphs; drinking vessels such as kraters or vase dedications such as aryballoi and lekythoi containing aromatic oils; and inscriptions on stone or graffiti on vases. The same type of material was found in the Cave of Daphni, [11] the Cave of Eleusis, [12] the Cave of the Lion on Mount Hymettos [13] and the Schisto Cave at Keratsini, [14] which suggests that these caves were also dedicated to the cult of Pan and/or the Nymphs, despite the lack of inscriptions confirming this assertion (fig. 1).

In all these cave sanctuaries, there is a strong decline in the quantity of archaeological material in the Hellenistic period, and a complete lack of material for at least part of the Imperial period. Alone, the lack of material from a specific period in a sanctuary is insufficient proof of abandonment. [15] Archeological sites are often disturbed by post-depositional events, and small, rural sanctuaries such as these Attic caves are often known exclusively through old and very brief excavation campaigns. However, the similar chronological profile of these eight caves can hardly be a coincidence, and it is likely that they were all visited less and less frequently in the Hellenistic period, and then almost completely abandoned by the Imperial period.

While ceramics from the Imperial period are completely lacking according to old excavation reports, recent research in some caves has brought to light a number of sherds from plain-ware ceramics. Some ceramics, lamps and coins from the Imperial period have been excavated in the Cave of Pan at Oinoe, [16] but this material is not precisely dated in the preliminary report and we cannot verify if there was a continuous occupation of the cave throughout the Imperial period. In the Schisto Cave at Keratsini, no material dating to the first three centuries of the Imperial period has been reported, but several fragments of coarse ware have not been dated. [17] In the Lion Cave on Mount Hymettos, recent excavations have brought to light amphorae, hydriai, oinochoai and cooking ware dated to the Hellenistic and Roman period. [18] It cannot be excluded that they were used for ritual dining, but such events would have been exceptional, and in all these caves we notice the complete absence of offerings dated to at least part of the Imperial period.

In the Classical period, the caves attracted a variety of visitors. In Vari, eight different inscriptions name a certain Archedamos, who described himself as a nympholept, a person seized by the Nymphs, but dedications were also made by a goatherd and by various women. [19] As suggested by the dedication of expensive marble reliefs, these caves constituted places of self-representation for men from the deme or even from neighboring demes. [20] The presence of a number of figurines representing women, but also of dedications such as loutrophoroi, indicates that the caves were also visited by women. [21]

There is no evidence of an implication of the polis or the demes in these sanctuaries. [22] The sanctuary of Pan and the Nymphs in Phyle is described by Menander as the “Nymphaion of the Phylasians” (τὸ νυμφαῖον […] | Φυλασίων), [23] but this passage does not necessarily mean that the cult was administered by the deme. First of all, it should be noted that the description of the sanctuary cannot be considered a faithful one. The cave is difficult to access — in Antiquity, people possibly had to cross a ravine to reach it [24] — while in the play, the shrine is in the vicinity of two houses. [25] Therefore, the mention of the Phylasians should also be analyzed in its literary context. One of the two major locations where the events of the play take place is Knemon’s house, where the character withdraws and refuses to open his door. By contrast, the sanctuary is described as a common place, open to all and where all characters gather at the end of the play. [26] This is what Menander may underline when he writes that the Nymphaion belongs to the Phylasians: here, the Phylasians probably represent the characters of the play. [27]

In the Vari Cave, a regulation from the second half of the fifth century instructs that intestines should be cleaned outside and the excrements washed off, [28] but no issuing body is named. Archedamos, though perhaps not the founder of the cult, [29] in a way appropriated the sanctuary, since he modified the cave by hand, planted a garden and built a house for the Nymphs. [30] It is therefore possible that he issued the regulation, [31] which is contemporary with the other inscriptions referring to him. Likewise, there is no evidence of a public administration of the cave of Oinoe. The mention of Pan in the sacrificial calendar of the Marathonian Tetrapolis, dated to the second quarter of the fourth century BCE, is an uncertain hypothesis. [32] In the year 61/60 BCE, three ephebes dedicated to Pan and the Nymphs a regulation prohibiting the wearing of colorful clothes, which was displayed at the entrance of the cave of Oinoe and which, unfortunately, is fragmentary. [33] It has been suggested that the ephebes did not have the authority to issue such a regulation and therefore they merely copied and rededicated an earlier ritual norm. [34] If so, it does not necessarily mean that this earlier regulation was issued by the deme, as suggested by a comparison with the case of Archedamos in Vari. Moreover, if we consider that the cave seems virtually abandoned in the first century BCE, there is no reason why the ephebes could not have issued this regulation themselves. They would have felt the need to remind the occasional visitors, such as shepherds and their flocks, [35] or other ephebes visiting a place associated with the battle of Marathon, [36] to abide by the purity rules that normally apply in sanctuaries, even if the cult was not the primary reason for their visit.

There is no indication that the cult in any of these caves was financed by public funding, or that sacrifices to Pan and the Nymphs were integrated into the sacrificial calendar of the demes. Because of this, it is impossible to pinpoint a precise moment of abandonment for these sanctuaries. In the Hellenistic period, they stopped attracting expensive dedications, and spontaneous visits by the local population became less and less frequent until they eventually stopped, or at least disappeared completely from the archaeological record. Can a sanctuary be considered abandoned if a shepherd or an ephebe makes a dedication there once a year or once every few years?

Deciding whether these sanctuaries were abandoned is even more problematic for the Imperial period. We notice a lack of archaeological material, or at least of material indicative of a cult, in all these caves. There is, however, an important exception. At least 24 so-called Alpha-Globule lamps from the first to second century CE were found in the Vari cave, [37] and at least 10 in the Phyle Cave. [38] In the Cave of Pan at Oinoe, three Alpha-Globule lamps and three contemporaneous lamps of a different type were known from the previous excavations, [39] and new research in the sanctuary brought to light new lamps, whose dates are not indicated in the preliminary report. [40]

Contemporaneous lamps from the Corycian Cave on Mount Parnassos, which was also apparently abandoned in the Hellenistic period, have been attributed to “tourists”. [41] At the beginning of his work On the Pythian Oracles, Plutarch writes that various characters who appear in the dialogue escorted a foreign visitor and showed him the various offerings in the Delphic sanctuary; then most of them went up to the Corycian cave with the foreign visitor, who is described as φιλοθεάμων, “fond of sights”. [42] Pausanias too visits the cave and writes that it is the “best worth seeing” (θέας ἄξιον μάλιστα) of all the caves he has visited. [43] He also mentions the Cave of Pan in Marathon, [44] probably the cave of Oinoe. In both cases, he seems more interested by the geological features of the caves than in their former use as sanctuaries. Even though he writes that the Corycian cave can mostly be visited without any additional light, visitors like him possibly brought lamps because they were unaware of the configuration of the cave or in order to explore its depths. The lamps would have been left behind as dedications to the pertinent deities. Lamps found in the caves of Vari, Phyle and Oinoe should perhaps be interpreted the same way. It is noteworthy that the date of the so-called Alpha-Globule lamps, between the middle of the first and the end of the second century CE, corresponds to the period of the Second Sophistic, when “intellectual pilgrimage” to old cult sites was widespread among the elites. [45] Whether such visits could be seen as an indication of a continuity in the cult is debatable.

In the Cave of Eleusis, a great number of figurines representing either Pan or women belong to the third and, to a lesser extent, the fourth century CE. [46] This material has been taken as an indication of a continuity in visits to the cave from the Classical to the Imperial period. [47] However, the catalogue of findings does not include a single object dated between the second century BCE and the third century CE, except for an Early Roman lamp. [48] Even though the cave was not completely excavated before it was destroyed in the 1950s, and all objects have not been dated precisely, it is possible that the cult was rather revived after a break of several centuries for unknown reasons.

The same can be said about the Cave of Phyle on Mount Parnes, where twelve inscriptions dated between the second and the fourth centuries CE indicate a revival of the cult. Three of the inscriptions are dedications, by artists or craftsmen, of images of Pan or of bronze vessels for libation and the burning of incense. [49] Other inscriptions were engraved by prominent men. One of them is “the famous (κλεινὸς) Thessalikos”, who is otherwise unknown. [50] One of the last dedications, in the fourth century, was made by Nikagoras, the son of a former dadouch from Eleusis, who comes from a well-known family with ties to the emperor Constantine. [51] Several of the inscriptions are in verse. This includes a dedication by three Thracian men, one of whom composed the verses to celebrate the glory of Pan. [52] According to this inscription, the dedication was set up “in the solitary mountains” (ἐν οἰοπόλοις ο[ὔρεσι]). Nikagoras also speaks of a difficult ascent ([χαλεπὴν μετ’ ἀν]αιβασίην, if the restoration is correct) and apparently visited the cave eleven or twelve times ([δωδ]ε̣κάκις). A man called Eusophis also climbed to the cave six times. [53] Another text is a topos-inscription: “So-and-so was here”. [54] Inscriptions only recording names in the nominative should perhaps be interpreted in the same way. [55]

It is noteworthy that no archaeological material from this period has been found in the sanctuary — even though we know from the inscriptions that at least some objects, such as bronze vessels, were dedicated. This suggests that there was no widespread use of the cave as a cult place in this period, but that it was rather a place of pilgrimage and self-representation for a few members of the elite, who might have seen the difficult ascent to the cave as a challenge. In the same period, Aelian, who mentions the Phylasian sanctuary of Pan in a rewriting of Menander’s Dyskolos, may be a witness to an intellectual interest in the cave. [56]

In Eleusis, after the (re)use of the cave in the third and fourth centuries CE, the site was probably abandoned, and no Late Antique lamps have been found. In the Phyle Cave, it is likely that the earliest Late Antique lamps, which are dated to the second quarter of the fourth century, are later than the latest fourth-century inscriptions, whose date cannot be determined more precisely. If so, why did the lamps replace the inscriptions? Should we assume a continuity or a change in the use of the cave? A closer look at these lamps is necessary in order to answer this question.

Around 2,000 lamps dated between the fourth century and the sixth or even the seventh century CE were found in the Vari Cave. [57] About 1,000–2,000 Late Antique lamps were also found in the Cave of Phyle, hence its modern name of Lychnospilia, “the Cave of the Lamps”. [58] In the Cave of Pan at Oinoe, at least 8 lamps dated between the fourth and the sixth centuries CE were found. [59] Three lamps from the fourth and fifth centuries are also reported among the finds of the recent excavations in the Schisto Cave at Keratsini. [60] With the exception of some coins and the occasional sherds of plain ware, [61] lamps constitute the only Late Antique material from these caves.

Almost all these lamps, if not all of them, bear signs of use. In Vari, olive stones were found in many lamps, showing that olive oil was used to light them, while part of the wick was preserved in another lamp. [62] In all of the caves, a majority of the lamps were found intact. Therefore, they cannot be compared to the Cave of Zeus on Mount Ida, in Crete, where a great number of lamps from the Imperial period were found broken into small pieces, suggesting that they were set alight and intentionally thrown into the cave as part of a ritual. [63] In Attica, the lamps were probably used to light the caves. However, it is striking that in both the Phyle Cave and the Vari Cave very few lamps were found from the Classical period, when the cult was at its peak. In fact, in Vari the Classical inscriptions are concentrated in the well-lit part of the cave and can be read without a lamp. [64] Archedamos the Nympholept has been described as a master of light because of the way he took natural light into account when he modified the cave. [65] Before reaching the main room, visitors would have passed through the Western part, the darkest of the cave, but in the Classical period they did not use lamps to light their way. Perhaps this passage through darkness was intended as a part of the worshipper’s experience. [66] In the Phyle Cave, only the front third of the cave was excavated. [67] It remains unknown whether the deepest part of the cave, which is almost completely dark, was also visited in the Classical period. However, all the inscriptions dated between the second and the fourth century CE were placed either outside the cave or near its entrance, and thus could probably be read without artificial light. This contrasts with the Cave of Pan at Oinoe, where most of the material found in the recent excavations, including some unpublished inscriptions, comes from the deepest part of the cave, which is deprived of natural light. [68] There, more than 30 Classical and Hellenistic lamps have been found, while only 8 Late Antique lamps are reported. [69]

If the caves of Vari and Phyle could be visited without artificial light, why were so many lamps deposited in Late Antiquity? In Vari, some lamps were found in the deepest part of the cave, where ancient finds are otherwise absent, but other lamps come from the most well-lit area. [70] In Phyle, all lamps were found in the front third of the cave, the only part that has been excavated. [71] There is no indication that the deepest part of the caves was preferred in Late Antiquity. However, it should be noted that the stratigraphy of all the caves under consideration has been greatly disturbed, particularly by the movement of water and by flocks taking shelter.

In fact, to understand the use of lamps, we must take into account the fact that caves are not the only places where a large number of Late Antique lamps have been found. In particular, fragments of about 120 lamps from the fourth and fifth centuries were found in the sanctuary of Zeus on top of Mount Hymettos. The sanctuary was mainly used in the Geometric and Archaic periods. A few finds indicate sporadic use until the first century BCE; then the sanctuary seems to have been completely abandoned until Late Antiquity. [72] Given that the sanctuary of Zeus is an open-air sanctuary, [73] which means that lamps would not have been necessary in broad daylight, it is likely that the lamps were used at night. That said, its history is comparable to that of the caves, and a common explanation should be sought for the presence of such a great number of lamps.

Throughout Greek Antiquity, oil lamps were the main lighting device used in closed spaces, especially at home. However, in large and relatively well-ventilated spaces like caves, it was also possible to use torches, though they burn out quickly and produce smoke, [74] or to light fires. In the Vari cave, at the beginning of the twentieth century, archaeologists saw shepherds burning bunches of thyme, and in some parts of the cave, the walls have been darkened by smoke. [75] In Phyle, all the archaeological material from the Classical period to Late Antiquity was found within a layer of ashes, which suggests that fires were also lit inside the cave. [76] Lamps are much less efficient than torches or fires. Their light mainly spreads in a horizontal direction, and only has a range of about 3 ft (1 m). [77] How much light they produce depends on the number of lamps lit at the same time. In the caves of Vari and Phyle, where more than 1,000 pieces were found, it would have been possible to set alight a great number of lamps at the same time. However, it is not at all certain that this is what actually happened. If, each year, ten persons brought a lamp each and left it in the cave, the number of lamps would have reached 2,000 in two centuries. Unfortunately, the dating of ancient lamps is not precise enough to allow us to determine whether a few lamps were deposited regularly over a long period of time or a large number of them were deposited on a few occasions.

In the Cave of Oinoe, where more than thirty Classical and Hellenistic lamps were found, one particular object can help us understand why lamps were sometimes preferred over torches to light the deepest part of the cave. A lamp from the end of the third century BCE bears the inscription Ἱερός, “sacred”, suggesting that it was not only used to light the cave but that it was also dedicated to the deities of the sanctuary. [78] Moreover, the flickering light of the lamps that would have left most of the cave in the dark would have contributed to the worshipper’s particular sensory experience [79] and perhaps even led to the altered state of consciousness that Archedamos called “nympholepsy”. Were lamps chosen for the same reasons during Late Antiquity?

If the lamps were used during religious rituals and perhaps deposited as offerings, who visited the caves in Late Antiquity and which deity/deities were these offerings addressed to? According to some scholars, lamps are evidence of a Christian use of the cave. [80] A few elements could support such an interpretation. Most of the lamps are undecorated or only show ornamental motifs, but a few bear representations on the disk. Some fourth-century lamps represent gods such as Athena and Pan, but Christian motifs appear in the fifth century. [81] Other traces of a Christian presence were found in the Vari Cave: a cross engraved in three places, including the relief depicting Archedamos; a finger ring decorated with the symbol of the cross; and the mutilation of the faces of the figures in the reliefs representing the Nymphs and Pan. [82] The same damage to the reliefs can be observed in the Phyle Cave. [83] However, this Christian presence cannot be dated and it is not certain that it is contemporary with the use of lamps in the cave. As for the lamps themselves, recent studies have stressed that their iconography does not always determine their usage: [84] Christian lamps were not used exclusively by Christians, especially given the growing availability of lamps with Christian motifs from the fifth century on. In fact, the Christian presence in the Vari Cave, and especially in the Phyle Cave, is particularly discreet. If the lamps were deposited in the Attic caves by Christians, we would expect more Christian symbols and inscriptions, which are completely lacking in Phyle. For the same reason, it seems unlikely that the caves were shared by various religious groups, including the Christians. [85]

Could a comparison with the so-called “Fountain of Lamps” in Corinth suggest that “magical” rituals took place in the Attic caves? In this underground bath and fountain that had fallen out of use by the fourth century CE, about 4,000 lamps dated between the fourth and sixth century were found, whole or broken. Some lamps bear inscriptions that are usually interpreted as “magical”, and four lead tablets, probably curses, were also found among this material. [86] Excavations in the fountain of Anna Perenna in Rome also brought to light 74 lamps from the fourth and fifth centuries CE, as well as “magical” objects, including 26 lead tablets, six of which were inserted inside a lamp. [87] Three quarters of the lamps were unused, in accordance with the prescriptions found in the Papyri Graecae Magicae for magical rituals involving lamps. [88] It has been suggested that in Late Antiquity, Nymphs associated with fountains were thought of as demons, and, therefore, that “handing over lamps to watery Nymphs could only signify handing over souls to evil female demons.” [89] Likewise, should we suppose that the Attic caves that were formerly dedicated to the Nymphs, and where water was flowing or dripping, became the focus of magical ritual in Late Antiquity? Such an interpretation seems unlikely. First of all, not a single magical text was found in these caves. Moreover, all lamps bear signs of use, unlike the lamps found in the fountain of Anna Perenna. But more importantly, some sanctuaries where Late Antique lamps were found, such as the sanctuary of Zeus on Mount Hymettos, were not dedicated to the Nymphs.

Another hypothesis is that the caves were visited by Neoplatonists who worshipped the traditional deities. [90] Porphyry, who wrote On the Cave of the Nymphs in the Odyssey, [91] and, according to Marinus, Proclus, who headed the Platonic school in the fifth century, often visited old cult sites to honor the gods, especially Pan. [92] It is therefore possible that Neoplatonists had a special interest in Pan and the Nymphs, and visited their caves. However, when compared to the much smaller number of lamps that may indicate intellectual pilgrimages during the Second Sophistic, or the limited evidence of elite visits to the Cave of Phyle from the second to the fourth century CE, it is unlikely that the presence of Neoplatonists alone can account for the thousands of lamps found in Vari and Phyle.

All these interpretations take for granted that the lamps had a religious significance, but the absence of material characteristic of sanctuaries is striking. Attic caves were used for various reasons in Antiquity: they were shelters for shepherds, storage places, water sources and even dumps, even if there is no evidence of long-term habitation in any Attic cave in the historical period. [93] None of the identified uses explains the presence of such a concentrated number of lamps. It is also important to note that a number of Late Antique lamps were found in places that were not sanctuaries. On the western peak of Mt. Kerata, a watchtower that had apparently been abandoned for centuries, the only materials dated later than the Hellenistic period were at least 21 fragments of lamps from the fifth and sixth century CE, three of which bear representations of Athena. [94] Other abandoned watchtowers also yielded some Late Antique lamps. [95] Near the Vari Cave, a country house that had been abandoned since the third century BCE yielded five lamps, two lamp fragments, a jug and a storage amphora from Late Antiquity. [96] More than 60 Late Antique lamps were also found in a mine gallery in Thorikos, which showed no sign of use in the Imperial period. [97] It is unclear if all these cases should be attributed to the same phenomenon. For example, the lamps in Thorikos have been interpreted as evidence of a reuse of the mine in Late Antiquity. [98]

Whether or not we take at least some of these non-religious sites into account for the interpretation of the lamps, all of them have one thing in common: they are all located in remote places and/or in abandoned sites. The Phyle Cave is inconspicuous, [99] and inscriptions stress how difficult the ascent was. From the site of Vari, the probable location of the deme of Anagyrous, visitors to the Vari cave would first have headed to the isolated country house, then would have reached the cave after another ten-minute climb up the Hymettos. The cave itself was accessible through a large hole in the ground, and therefore could not be seen from afar. [100] The Marathon Cave is fairly inconspicuous, too. [101] It is located on the same hill as the acropolis of the deme, but Pliny suggests that the deme center was abandoned in his days, [102] and it is unclear if it was still occupied or reoccupied in Late Antiquity. The sanctuary of Zeus on Mount Hymettos was located on top of the mountain, in a natural depression that hid it from view. [103] Late-Antique lamps were also found in the abandoned sanctuary of Zeus on the summit of Mount Parnes, either inside or in front of a cave. [104] At least ten lamps from the fifth and sixth centuries were found in the abandoned sanctuary of Varkiza, which was located in an isolated area probably visited only by shepherds. [105] By contrast, there were no lamps in the Cave of Eleusis, located near the acropolis of the deme, or in the Cave of Daphni, which was visible from the Sacred Way.

As suggested above, the lamps were probably used at night. Why did people go to remote and inconspicuous places at night between the fourth and the sixth or seventh century CE? The obvious answer is that they were polytheists who had to hide in order to continue to honor their traditional gods. [106] The chronology of the coins found in Vari may support such a hypothesis. Of the 147 coins, only one is Hellenistic, and 82 are from the fourth century CE. [107] The 64 unidentified coins could belong to the fifth and sixth centuries, since coins from that period are often difficult to identify because of their poor quality and small size. [108] Among the fourth-century coins, only one belongs to Constantine the Great. More than half (46), belong to his successor Constantius II (337–361 CE). There is then a fairly even distribution of coins until the reign of Arcadius between 395 and 408 CE (34 in total).

The large number of coins under Constantius II does not necessarily mean that there was a peak in the number of visits to the cave during his reign. In the excavations of the Athenian Agora, the number of coins of Constantius II surpasses by far the number of coins of any other emperor. [109] Even though coins can circulate for a long time and only indicate a terminus post quem, it is striking that they start appearing in the caves precisely under the first emperor to ban sacrifices on pain of death, who decreed that sanctuaries should be closed. [110] After this first ban, traditional cults were reauthorized by emperor Julian, then tolerated by his successors, until they were banned again under Theodosius in 391/392. [111] What exactly these legislations forbade is a matter of debate, and the bans did not really put an end to polytheistic cults, [112] especially in Attica. [113] However, as Libanius’ writings suggest, [114] perceived danger may have encouraged some persons to hide in caves in order to honor the old gods. Both in Vari and in Phyle, the presence of fourth-century material is relatively limited, and the presence of lamps increases in the fifth century, when the traditional cults became less and less tolerated.

Such an interpretation would also explain the absence of other types of material traditionally associated with the cult of Pan and the Nymphs. A lamp and its light could be deposited as an offering to a deity, [115] but it was still an everyday object and there was no danger in carrying it around. The presence of a large number of Late Antique coins in the Vari Cave, as compared to the absence of Classical coins, is also easier to explain if we consider them as offerings. [116] In fact, some Christian authors suggest that this is precisely what those who wanted to continue honoring the traditional gods did: they went to remote places, lit lamps, threw coins into springs, offered incense, tied ribbons to trees, all actions that involve everyday objects and not objects that were specifically made to be deposited as offerings. [117] If remote places were chosen to honor the traditional gods, this may explain why some lamps were also found in non-religious sites such as abandoned watchtowers: ruins were a good hiding place, even if sanctuaries were preferred. If so, it is not certain that the people who brought lamps to the cave dedicated them to Pan and the Nymphs specifically; they may have looked for a safe place to honor all their traditional gods at the same time.

It is likely that Late Antique lamps deposited in abandoned sanctuaries such as caves dedicated to Pan and the Nymphs or mountain-top sanctuaries of Zeus are indeed evidence of cult practice. However, they can only be understood in the context of Late Antiquity. The absence of other types of material usually found in sanctuaries is not surprising because worshippers would have avoided dedicating objects that could immediately be recognized as offerings to the traditional gods. By contrast, lamps from earlier periods cannot be considered sufficient evidence of cult continuity because the lack of other types of material characteristic of sanctuaries needs to be explained.

Attic caves also allow us to reflect on the memory of abandoned cult places. If the location of these caves was forgotten by all but the shepherds who used them for shelter, the presence of old inscriptions and reliefs would still have allowed the visitors to recognize that the caves were former sanctuaries dedicated to Pan and the Nymphs. It is also possible that the location of the cult sites was never forgotten, even when they fell into disuse. Whatever the case, the dedication of a regulation by ephebes in the Cave of Oinoe, the Alpha-Globule lamps and the inscriptions engraved in the Phyle Cave in the Imperial period indicate that there was an elite interest in these abandoned cult places. These rare visitors were probably attracted by the geological features of the caves, such as the stalactites and the stalagmites or the history of the sanctuary, or they considered the difficulty of access to the cave as a challenge, but they still saw the caves as sacred places and probably made dedications to their deities. As this example shows, it is not always possible to draw a clear distinction between an abandoned sanctuary and a sanctuary that remained in continuous use.