Lamellae nos. 1–7 were recovered from graves in the extensive cemetery to the north-northwest of Eleutherna, the city’s north entrance, but the exact location of their discovery is unknown. Theodoros Triphyllis, Consul of Austria-Hungary in Rethymno came to acquire nos. 1–3 and 7 and showed them to André Joubin, [1] John L. Myres, [2] and Federico Halbherr  [3] before his collection of antiquities ended up in the National Archaeological Museum. [4] He told these men that the lamellae came from graves in Eleutherna, assuring them further that they came “from the same grave,” [5] a rather unlikely piece of information. [6] Halbherr, who visited the site and inquired about the lamellae’s provenance, was told by locals that they came from the large cemetery at the north entrance of Eleutherna which reaches the modern villages Lagká (Λαγκά, ‘Ravine’) and Alfá (Αλφά, ‘Limestone’).

Thus Margarita Guarducci, drawing on Halbherr’s notes, included four lamellae in the epigraphical dossier of Eleutherna, because they were found non longe ab Eleutherna et quid loco Alphá (ita Halbherr, in schedis; qui de hoc a nonnullis loci illius incolis se certiorem factum esse adfirmat), in sepulcreto aliquot … [7] The lamella in Herakleion Museum was published by Guarducci among the inscriptions of Loci Incerti, because in Cretae regione Mylopótamos appellata reperta est (e quo potissimum loco prodierit non constat) … [8] Chances are, however, that as Guarducci proposed later, [9] this lamella (no. 4 above) came from the same cemetery as nos. 1–3. In fact, it is probable that nos. 5 and 6 in the Hélène Stathatos Collection also came from this cemetery, since they were all found, according to the seller, “in graves near Eleutherna.” [10]

Although no systematic excavations have been undertaken and sufficient evidence is therefore lacking, the extensive Hellenistic-Roman cemetery in the sites Mnemata (Figures 24–27, pages 82–84) and Agia Elessa (Figures 29–33, pages 85–87) appears to be the most likely candidate for the provenance of the seven Cretan epistomia (see Acknowledgements above).

Some years ago, a brief rescue excavation revealed a Roman bath (Figure 28, page 84) in close proximity to Agia Elessa. [11] The site is called Mnemata (Μνήματα, ‘Graves’; Figure 24, page 82), because the graves to the south and east of the village Alfá are still visible (Figures 25–27, pages 83–84), whereas Agia Elessa (Ἁγία Ἐλέσα or Ἐλεοῦσα, ‘Holy Elessa or Virgin Mary the Compassionate or Saint Helen and Constantine’; Figure 29, page 85) is nothing more than a Roman chamber-tomb with larnakes that was later converted into a small church (Figures 30–33, pages 85–87). [12] This fact may account for the locals’ insistence, upon being questioned by Halbherr, that the lamellae came from the same grave, i.e. from a Roman chamber tomb with more than one larnax (Figures 30 and 32–33, pages 85–87). A similar example may be observed in ancient Lappa (modern Argyroupolis; see map, opposite page 1). In the Roman cemetery to the east of the city, one chamber-tomb became the small church of Agies Pente Parthenes (Ἅγιες Πέντε Παρθένες, ‘Holy Five Virgins’; Figures 34–37, pages 87–89) and another tomb was converted into the Agia Elessa church (Figures 38–39, pages 89–90). [13]

The sites Mnemata (Figures 24–28, pages 82–84) and Agia Elessa (Figures 29–33, pages 85–87) to the north-northwest of Eleutherna are located on hilly terrain approximately 200 m above sea-level, on the borders of the modern villages Lagká and Alfá. [14] The graves are cut on top of a hill. The west hillside is very steep due to erosion (Figures 26–27, 29, pages 83–85). Thanassis Kalpaxis and Niki Tsatsaki have studied the sites and the chance finds from some of the looted graves dated from the late fourth to the second centuries BCE, and they have suggested that this cemetery was probably used by the inhabitants of the Nesi hill in Eleutherna, whose remains are dated within the same period. [15] This extensive cemetery lies within the wider area of Eleutherna, the ancient city at the northern foothills of Mount Ida. As the excavations by the Department of History and Archaeology of the University of Crete have shown, [16] Eleutherna’s settlement pattern emerges as one of many ‘neighborhoods’ at some distance from each other, what van Effenterre has called “un habitat polynucléaire.” [17] Nicholas Stampolidis has convincingly argued that the graves to the east and west of the two streams of Eleutherna, which eventually flow north into the river Geropótamos, are located at Eleutherna’s natural passages from the north. [18] The village Lagká is located approximately 2 km to the north-northwest of Eleutherna and the village Alfá approximately 4 km, the latter roughly at the midpoint of the approximately 10 km distance from Eleutherna to the north shore (see map, opposite page 1).

Lamellae nos. 8–12 come to us from rescue excavations which have revealed part of the Hellenistic-Roman cemetery in Sfakaki on the north shore. [19] It seems certain that this cemetery belonged to the settlement that has been excavated in the villages Stavromenos and Chamalevri. It is therefore reasonable to assume that this settlement, most probably within the wider area of Eleutherna, would or should bear some similarities to Eleutherna in terms of political, social, economic, and religious matters. From the religious dimension, at least, we find support for this ostensible overlap in the fact that the nine incised epistomia form a distinct group (albeit not a homogeneous one) and consequently must be studied as texts related to, even if not actually produced in, the city of Eleutherna.

The exact location, however, of the ancient cities on the north shore of Crete from Rhithymna to Herakleion remains vague. It is difficult to determine the territories of Cretan cities in antiquity, as they shifted from period to period on account of various economic, social, and political changes. [20] Nevertheless, we can be certain that Eleutherna’s territory extended to the north shore in the modern villages of Stavromenos, Chamalevri, Sfakaki, and Panormo, where one or possibly two of the city’s harbors must have been. The ancient cities(?) Allaria, Arion (Ariaioi), Pantomatrion, Panormos, and the site Dion Akron are variously associated with Eleutherna and with the north shore of this part of the island (see map, opposite page 1).

Dion Akron, a site mentioned in a text prescribing that Eleuthernaean dromeis serve in this promontory-outpost(?), [21] has been placed on modern Cape Stavros, west of Sisai, probably an ancient dependent city. [22] Stampolidis has suggested persuasively that it should instead be placed somewhere between modern Panormo and Bali (ancient Astale), Dion Akron probably being Eleutherna’s east border with Axos. [23]

Stylianos Alexiou places Panormos in modern Agia Pelagia, ancient Apollonia, and locates Arion in Stavromenos/Chamalevri, and Pantomatrion in modern Panormo (Castel Mylopotamo), although this does not exclude the possibility of an Eleuthernaean harbor in the area with some other name, presently unknown. [24] In Panormo’s vicinity, the Melidoni Cave, the Tallaeum Antrum dedicated to Hermes in the Roman period, may have served as a border sanctuary between Eleutherna and Axos, although its exact role may have changed from one period to the next. From the Roman period onwards, it appears that caves regain their attractiveness as potential sites for cult places and sanctuaries. [25] During the Imperial period, however, Eleutherna was privileged over Axos by the emperors, [26] and her territory may have been extended again to include the Melidoni Cave—although, under Roman rule, what such control meant is unclear.

Other equally convincing possibilities for Pantomatrion, Allaria, and Arion (Ariaioi) include the modern villages of Sfakaki, Chamalevri, and Stavromenos, all of which lie 8–10 km east of Rethymno, [27] the area where one of Eleutherna’s ports was probably located, and the area which most probably served as the city’s western border with Rhithymna. We do not have sufficient evidence for determining the eastern border of Rhithymna’s territory, nor can we ascertain whether or not this territory was fixed throughout the centuries. Guarducci included in the epigraphical dossier of Rhithymna a small number of inscriptions from this region. [28] Albeit of no help in identifying the ancient settlement, at present these texts comprise the area’s epigraphical dossier, together with the two incised epistomia from Sfakaki and a new inscription from Chamalevri that corroborates the existence of a hieron and most probably a city. [29] The continuing excavations by the 25th Ephorate are gradually bringing to light extensive settlements in this region, [30] but unfortunately the evidence is still too inconclusive to allow for identification with any known ancient settlement.

Nonetheless, the topography of the graves within the Hellenistic-Roman cemetery in Sfakaki, although the cemetery’s limits remain unknown, does not support the idea that a special burial place was set aside for those deceased bearing lamellae and epistomia. The idea, unanimously accepted, that within a cemetery there may have been reserved areas for particular burials, appeared after the publication of an inscription from Kyme. This text, however, inscribed in five lines on a block from a large chamber-tomb and dated to ca. 450 BCE, prohibits burial inside the tomb of those not initiated in the Bacchic mysteries, [31] a case similar, if not identical, to the ἀραὶ ἐπιτύμβιοι. [32] The text does not imply a special area within the cemetery designated for burial of Bacchic initiates, but forbids burial inside the specific chamber-tomb on whose block it was inscribed. If another bebaccheumenos passed away, presumably the chamber-tomb might be re-opened for the new burial, but the main purport was that this tomb should not be re-opened and violated for the interment of just anybody. In turn, this should not lead to the conclusion that all prohibitions and curses against violation of the grave imply that the deceased was an initiate; only that the deceased with ἀραὶ ἐπιτύμβιοι and his family for some reason were strongly against re-opening of the grave or its violation.

The idea that a special burial place was set aside is not evident in the partly excavated cemetery in Sfakaki. Of the fifty–six, only five graves contained an epistomion (nos. 8–12): Grave 4 (no. 11 above) is immediately to the west of Grave 1 (no. 8 above), and Graves 9 (no. 10 above) and 20 (no. 12 above) are in relative proximity, but Grave I (no. 9 above) is some distance away. At present, it does not appear that a specific area within the excavated part of the cemetery was reserved for mystai, [33] as is also the case in the Amphipolis cemetery, where of more than 1600 graves excavated, only one contained an engraved lamella (D4). [34] If there were ever a choice for a burial plot within a cemetery, or if designated areas ever existed, the topographical indications are inconclusive; they do suggest, however, that most probably such choices may not have been based on a distinction or aggregation according to wealth, religious conviction, and so forth.

The letters on the small and extremely thin gold foils were incised by a small, sharp instrument, without the engraver tearing the foils. The letters in no. 8 above are an exception, as they seem to have been pressed. It is not certain whether these lamellae were first cut to their present size and then incised, or whether the texts were engraved first in longer foils and then cut with scissors accordingly, perhaps for mass-production. [35] At least some unincised epistomia appear to have been cut with scissors, not always successfully, after they have been decorated. [36] Nor can the possibility be ruled out that some of these gold foils originally had some other use and that, when the occasion arose, they were cut and incised or vice versa.

The need to keep the lamella steady for incision and for the laying-out of the text most likely explains the indention and the spaces left vacant in all the lamellae and can further offer some hints about the lamellae’s preparation. [37] Nos. 4, 5, and perhaps 6 were probably incised in their present size. The text of no. 4 is etched following the mouth-shape of the foil without care for a neat layout. In nos. 5 and 6 on the other hand, there seems to have been a miscalculation regarding the text to be engraved relative to the surface area available on the foils: in no. 5 the text is crammed on the upper two-thirds with enough space left empty below the text for one or two more lines; in no. 6 the same amount of letters as no. 5 fills in almost the entire surface, as the characters are slightly taller, and the lines are meandering.

The layout of nos. 1–2 (and to a lesser extent, no. 3) is careful and ordered, the lines are straight, and syllabification is respected. Symmetry in incising the letters from left to right in no. 7 is more important than word-division, whereas in no. 8, the two words are centered. The text in no. 9 is laid out differently, although from the middle of line 4 onwards, the letters become larger in order to fill the space which otherwise would have been left empty. Even so, the last line is incised in such a way that the lamella gives the impression that it had been cut after incision, as if one had cut(?) the foil carelessly and without knowing where exactly it should be cut; such an inference can perhaps explain the text’s abrupt break off.

The lettering (see nos. 1–9 above), with minor idiosyncrasies and sometimes with different shapes within the same text, is basically similar in all lamellae, except in no. 8. In nos. 1, 2, 5, 7, and 8, the style of the letters is very careful and ordered, with very few mistakes—the work of an experienced, if not a professional(?) engraver. In nos. 3, 4, 6, and 9, on the other hand, the lettering is sloppy and offhand, perhaps the work of an amateur. Comparetti, based on an analysis of lettering, layout, and orthography, suggested that lamellae nos. 1 and 2 were apparently incised by the same engraver and that no. 7 is very similar to nos. 1–2, a conceivable case accepted by Guarducci. [38] These types of assessments, however, are hard to prove and the differences in the style of the letters, although minute (see nos. 1–3, 7 above), do not support such attributions. Engraving objects as small as lamellae was difficult and demanding, and resulted in numerous shapes for individual letters (see the section “Chronology”).

In terms of orthography and dialectic forms, the texts appear consistent and follow the rules of the Cretan-Doric psilotic dialect: [39] the infinitive ending in -εν in nos. 1–6 and 9 πιε̑ν, and in no. 7 χαίρεν; [40] the 1st person singular of ἠμί in nos. 1–6 and 9; [41] in nos. 1–6, the genitive singular in ω = ου, but also in the beginning of Ὠρανῶ and in the interrogative πῶ = πόθεν; [42] the shapes Σ or Ζ for -σσ- in the word κυφάριζος and in the 2nd person singular ἐζί of εἰμί (the shape Σ, according to Guarducci, is perhaps a relic of the archaic/classical tsade (sampi), turned 90 degrees counterclockwise). [43] In nos. 1–3, 5–6, and 9, the aspiration of the word κυφάριζος is not uncommon, nor are the two forms for Persephone’s name: Φερσοπόνη in no. 7, with metathesis of aspiration, is a rarer form than Φερσεφόνη in no. 8 with two aspirates, the normal spelling (together with the Attic form Φερρέφαττα) in pre-Roman Attic inscriptions. [44]

The orthographic variations of the texts in nos. 1–6 do not present major difficulties; perhaps Verdelis (1953–1954:57) is correct in describing the engravers of nos. 5–6 as having only a mediocre knowledge of grammatical rules. The only noteworthy case of variation involves the epithet for the spring: in nos. 1–2 and 5, it is αἰειρόω, and in nos. 3 and 6 (the more problematic texts in terms of letter-style and engraving) αἰενάω. Olivieri (1915:15) inferred that αἰέναος must be the older form whose glosseme αἰείροος became the more comprehensible one, but Guarducci (IC II, p. 170) suggested that either one may have been in the archetype, if such a text ever existed. Zuntz (1971:363) noted the perfect suitability of αἰέναος, preserved in the faulty texts, as opposed to the more problematic αἰείροος.

Although the text on lamella no. 9 is similar to nos. 4–6 in terms of orthography, amateurism, and dialectic forms, it nevertheless stands out in comparison to the others because it presents significant dialectic and textual divergences. Instead of using either one of the epithets αἰέναος or αἰείροος, it employs a new and unexpected word <Σ>αύρου or Αὔρου for the spring, and as a consequence strongly suggests that the epithet in no. 4 ΑIΓIΔΔΩ deserves more thought before emendation. Because of ignorance (or error?), the location of the spring in no. 9 changes from right to left, and the epic τῆ followed by the nominative becomes the article of the noun in the genitive. In this text, both endings of the genitive singular -ου and -ω are present in <Σ>αύρου, κυφαρίζω, <Ο>ὐρανῶ; and instead of the epic ἐζί (ἐσσί) used in the other texts for the 2nd person singular of the verb εἰμί, the classical form εἶ is employed. If the verb’s change may be justified as a later simplification of the less intelligible and more archaizing form ἐζί, the spring of auros is a unique instance (it is mentioned only once, in Theophrastos’ Historia plantarum 3.3.4, in reference to the Idaean Cave), and perhaps a reference that may even antedate the form αἰέναος, which Olivieri has argued is the more pristine form (see the section “The Cretan Context of the Cretan Epistomia ”). Lastly, the latter part of the text of no. 9 is another unique instance and is thus equally difficult to account for: if not a hodgepodge, it appears to prompt further questions and answers regarding the stichomythia and the currently known narrative scenes (see the section “The Cretan Texts in the Context of a Ritual and a Hieros Logos ”).

The brief texts on nos. 7–8 are not metrical, but it is uncertain whether the two words were completely without rhythm, if any rhythm were intended. The layout of the remaining texts nos. 1–6 and 9 creates rather than resolves metrical problems. In nos. 1–3 and 5, the text’s layout appears to have been prearranged according to its metrical form, which presumably was recognized as such by the engraver and perhaps also by the buyer/recipient who placed the order. Their knowledge, however, of the hexameter or of its rhythm, if any, was only nominal, as αἰειρόω in line 2 is not analyzable in dactyls. Any knowledge in these matters, then, was probably limited to the recognition of a line as metrical by its incised form on the foil. The text in nos. 4, 6, and 9 (the most problematic ones), however, testifies that such a prerequisite or demand was not necessary in certain cases and that the engraver was sometimes free to incise the text as he pleased.

Even so, and despite the mystery surrounding this process, the text which the engravers of all seven lamellae, except no. 9, strive to etch either from memory or by copying, is in its Cretan-Doric dialect the following:

Zuntz suggested that the text “evidently consists, in the main, of three hexameters which are, however, oddly expanded.” The third hexameter may have originally comprised the question, attested in the Homeric epics in its brief form τίς πόθεν εἶς, and line 4, which may have originally been without the expendable anyway υἱός (or παῖς), in light of the variation in nos. 4 and 9; hence the original must have been a perfect hexameter: τίς πόθεν εἶς; Γᾶς ἠμι καὶ Ὠρανῶ ἀστερόεντος. Zuntz also noted the perfect suitability to meter and context of αἰέναος, preserved in the faulty texts (nos. 4 and 6), in contrast to αἰειρόω, “a rare word not unsuitable in itself but ruinous to the metre—unless indeed one supposes Ionic (or Lesbian) αἶϊ to have been expelled by local αἰεί.” [45]

Interestingly, in the Homeric epics the question is never employed in an Underworld context (e.g. in the Nekyia). When asked, [46] and depending on the way it is answered, the question almost always involves actual or metaphorical death or danger. In the battlefield, Achilles asks Asteropaios this question and kills him (Iliad 21.150); Diomedes asks the question to Glaukos, who is saved, however, on account of the previous hospitality (Iliad 6.123). In scenes of hospitality, the person asked the question reveals his identity only in ideal circumstances, as in the meeting of Telemachos with Nestor (Odyssey 3.71). In all other circumstances, the person faces two choices: either to answer earnestly and face imminent danger, as Odysseus realizes, when he replies to Polyphemos (Odyssey 9.504); [47] or to answer with a Cretan tale (Odyssey 13.253–286, 14.191–359, 19.164–202). The same pattern is followed in two of the Homeric Hymns where the question occurs. In the theoxenia scene, Demeter is asked the question and replies with a Cretan tale (Homeric Hymn to Demeter 113); and the Cretans, after being overtaken by Apollo as a pirate, reveal their true identity (Homeric Hymn to Apollo 452), [48] but as a consequence, change it, by becoming priests of Apollo at Delphi (as their failed nostos , which amounts to death, indicates). In all of these instances, the inquirer is in a privileged position, whereas the fate of the person inquired is sealed by the way he chooses to answer. If he passes the test, then his earlier precarious position is overcome.

This epic strategy is adopted in the texts of the lamellae. The inquirer is in power, and the question-and-answer test, so familiar from the epic world, allows the mystes, by answering honestly and as taught during initiation, to acquire a new identity in the Underworld, to bypass death, and to begin a new life after death. The fact that the Homeric question in line 3 may be thematically insignificant perhaps led Zuntz to deny the possibility that line 3 may also be in “rhythmical prose”—his own well-chosen term used in discussing the symbola or synthemata of the initiate’s deification in the texts of group A. [49] And yet both questions in line 3 meet his criteria for “rhythmical prose”: they are parallel, and they have an equal number of syllables and accents (two long syllables and a short, ‒ ‒ ⏑, which may in fact be two palimbacchiacs, a Dionysiac[?] meter). [50] In addition, these ‘questions’ follow, in a more moderate way, the pattern of grammatical, phonetic, syntactic, and semantic parallelism specified in relation to the Pelinna text by Charles Segal and Calvert Watkins. Watkins has further suggested that “the strophic alternation of metrical long lines and non-metrical short lines closely recalls the liturgical pattern of the Vedic Asvamedha.” [51] For the metrical form in line 4, Gallavotti and Tessier have proposed a combination of a reizianum and an enhoplion, [52] but it seems that the line’s dactylic rhythm (actually five dactyls) harkens back to that in lines 1–2, especially if υἱός is read ⏑⏑ as in epic. [53]

The text in lamella no. 9 presents metrical difficulties in the initial foot of the first hexameter, and these problems only worsen in line 2 with the change of the spring’s epithet (Σαύρου/Αὔρου) and the spring’s location ἐπ᾽ ἀριστερά. Such difficulties are comparable to the ones in texts nos. 1–2, and 4–5, where the epithet αἰειρόω is employed.

These difficulties notwithstanding, the majority of these brief texts comprise two dactylic hexameters (1–2) with a spondee only in the first foot, a line in the ritual’s rhythmical prose ‒ ‒ ⏑, and a fourth line again in dactylic rhythm. The new, bacchiac rhythm in line 3 changes the dactylic hexameter and at the same time announces the change of speaker, who initiates the brief stichomythia, but also signals a change of epic expectations. [54] At the level of meter and epic conventions, the change in rhythm may also indicate the new way of achieving immortality. Whereas epic poetry cannot promote any way of transcending death besides kleos (through which a mortal can become a hero through the epic poetry itself), the texts on the lamellae introduce another way, alien to epic poetry, but all the while employing the tricks of the epic craft. The epic model of immortality, expressed in dactylic hexameters (lines 1–2), is undermined first by the bacchiac rhythm of line 3 (the inquirer) and then by the dactylic rhythm of line 4 (the initiate), both of which pronounce an alternative route to immortality through initiation into a mystery cult.

Based on lettering, orthography, and a comparison with the Italian lamellae, Comparetti accepted Halbherr’s dating of Cretan epistomia nos. 1–3 to the second century BCE, a date also accepted by Guarducci for no. 4. [55] By contrast, on the basis of a palaeographic study of the letter-forms and orthography of nos. 5–6 (which he compared to the previously published nos. 1–3 and 7), Verdelis proposed a date no later than the middle of the third century BCE for all seven texts, nos. 1–7. [56] Epistomia nos. 1–7 lacked any archaeological context; consequently, any effort at dating them had to be based on palaeographic criteria, which involve personal criteria as well. The result was the two different dates for the Cretan epistomia. Epistomia nos. 8–9, however, discovered during rescue excavations and thus less difficult to date because of other grave-goods, accentuate vividly the precariousness of strict reliance on palaeographic criteria. [57] If epistomion no. 8 were to be dated exclusively on palaeographic grounds, then the letter-forms would certainly indicate the third, if not the late fourth century BCE. And yet, the bronze coin and the other grave-goods offer a secure date for the grave between the last two decades of the first century BCE and the first half of the first century CE. In like manner, although the grave was found disturbed, epistomion no. 9 cannot have been placed inside the grave before the second or even the early first century BCE.

The chronological inconsistency between letter-forms and other grave-goods in the case of epistomion no. 8 raises an issue seldom touched upon in discussions of the incised gold lamellae and epistomia. Dating is most often not absolute but relative to various factors: e.g. the date of the artifacts’ manufacture may not always coincide with the period in which they were used as grave-goods, except of course for the exclusively entaphia-artifacts. In the case of epistomion no. 8, it appears that either the engraver wished to use more archaic-looking letter-forms, or, rather unlikely, the incision of the epistomion took place during the third century BCE, and it somehow came to the possession of the deceased, or his family and was employed for his burial sometime between 25 BCE and 40 CE. [58]

It must be emphasized that the dates for these texts incised on the Cretan epistomia indicate only the time when the epistomia were placed in the graves. As has been noted, those epistomia discovered in rescue excavations at Sfakaki seem to fall within the second or first centuries BCE and the first century CE (based on the graves’ typology and other grave-goods), whereas nos. 1–7 may be dated to the third-first centuries BCE. The texts themselves, however, may actually antedate even the third century BCE. The dialectic and metrical forms of the Cretan texts are usually referred to as ‘archaisms’ and/or local translations of the Ur-Text. This “aura of archaic sanctity,” as Zuntz described it, [59] is conveyed by all of the incised epistomia. More specifically, in the Cretan texts, the major factors in evaluating archaism are not only the lettering in no. 8 and the dialectic and orthographic forms in all nine texts (which, after all, may have been the original ones translated into a more acceptable Panhellenic epic Kunstsprache), but also the combination of the dactylic rhythm with the rhythmic prose of the two questions (and possibly more than two, as implies no. 9 above) that are preserved verbatim in nos. 1–6 and 9 (B3–8 and 12) and in the lamella from Thessaly (B9 and Figure 41, page 91). Furley and Bremer have indicated that in dating the inscribed hymn from Palaikastro (no. 16 above), three separate issues must be distinguished: the date of the inscription, the date of the hymn, and the date of the cult behind the hymn. [60] Likewise, these three different aspects should be distinguished in dating the epigram from Phaistos (no. 17 above), and in dating all of the incised lamellae and epistomia. In particular, the fact that the Cretan epistomia may be dated somewhere between the third century BCE and the first CE does not necessarily bespeak the date of the texts’ composition, or the date of the ritual behind the epistomia, which undoubtedly antedates the placement of the epistomia in the graves. How far back one should go in assigning a date to the text’s composition, or to the ritual’s appearance on the island (whether it be the third, fourth, or even fifth century BCE) cannot be determined.

All twelve epistomia from Eleutherna and Sfakaki, like the majority of the lamellae so far published, are paper-thin foils of gold. The only exception to this rule is one foil of silver from Poseidonia (D1). [61] There are also three cases where, instead of a gold foil, two gold coins (nos. 13–14 above; F8, F9), and a small gold disc are employed (F12). Gold, on account of its natural properties and qualities, became a symbol for eternity and life after death; the practice of depositing gold objects in graves is, as Zuntz keenly observed, “unlikely to have been a mere ostentation of riches, just as its opposite, the dark and heavy lead, was used to promote destruction and death. It does not, in fact, seem unreasonable to assume that the Orphic lamellae were consciously devised as a positive counterpart to the traditional defixiones.” [62] The texts on the gold lamellae are literally and metaphorically gold because of their valuable promise for access to a golden world, [63] and because “both the immortalizing gold and the ivy-leaf shape reinforce their message by the physical form of their medium,” as Segal aptly put it for the Pelinna lamella. [64]

Cretan epistomia nos. 1–3, 5–7, 9, and 11–12 are oblong, whereas nos. 4, 8, and 10 are ellipsoid, in the shape of the mouth, indicating their use as an epistomion. This word has become a terminus technicus at least among the majority of Greek archaeologists, who have no difficulty in identifying paper-thin gold foils as epistomia, [65] using as a definitive criterion the position of the foils inside the grave.

The custom, however, of covering the mouth or the whole face of the deceased did not start with the incised lamellae, nor did it end in late antiquity as nos. 18–25 above indicate. The epistomia date from the end of 5000 BCE until the second and third centuries CE, and their shapes are relatively few: oblong, rhomboid, ellipsoid in the shape of the mouth, and (very rarely) rectangular or triangular. The decorative motifs of the unincised epistomia, embossed or engraved, vary greatly. [66] The majority of these artifacts were thought to be jewelry, which formed part of the body’s kosmos, until Pierre Amandry suggested that the lamellae are descendants of the Mycenaean gold masks. [67] Instead of covering the whole face of the deceased, people gradually employed (for economic and perhaps for more practical reasons) smaller foils for the forehead, eyes, mouth, and ears. [68] Aikaterini Despoini (1998) has shown incontrovertibly, under the telling title Gold epistomia, [69] that, as the excavations of the cemetery at Sindos in Macedonia during 1980–82 and subsequent discoveries in graves throughout Macedonia testify, the paper-thin gold foils, which were found near the cranium or even on the chest of the skeleton where some of these may have slipped (and which were described earlier as diadems or pectorals), had been used as epistomia, mouth-bands. Shape and motif of these epistomia do not appear to be important factors, except that these mouth-bands, rectangular, oblong, or rhomboid, approximate the shape of the mouth. The drawing by Arnold von Salis (Figure 40, page 90) offers an idea of the way in which these epistomia were fastened behind the head with a string passing through holes on either end. [70] Despoini, however, suggested that in all probability they were sewed on the garment that eventually covered the head, but, as not all epistomia bear marks of a needle, they may have been simply put on the mouth or on the chest of the deceased. The grave-goods in some of these graves, which are published, do not delineate a recurrent pattern so as to substantiate a distinct burial custom for those deceased bearing an epistomion, incised or not. [71] As a result, the reasons for this custom or its presence in different areas and especially in different times cannot be ascertained, despite Lucian’s satire of this custom wherein the deceased is unable to open his mouth and speak, as his jaws are tied up by the cloth. [72] To state the obvious: people in different periods and in different areas felt the need to cover the face or, more specifically, the mouth of the deceased.

There is, however, a special category that deserves attention: the incised or unincised lamellae in the shape of certain leaves, a word which sometimes is also employed for the incised lamellae to describe thinness, regardless of shape. [73] Some of these leaves were used as epistomia, but not exclusively. Unfortunately, the excavators’ preliminary reports seldom provide detailed information regarding the exact findspot of these gold leaves or of their accompanying grave-goods (Tables 1–2). Even when such information is provided, few students have paid attention to the archaeological context of the lamellae, in order to understand shapes, material, usage, and texts. [74] It is therefore worthwhile to survey briefly the evidence for the grave-goods, where available, accompanying each of the nine incised leaves and each of the remaining thirty–five incised lamellae which have been published, in order to gain a general picture, especially on the issues of shape, context, and usage.

Only eight (D2, E3, F2, F4, F5, F6, F7, F11) incised gold leaves have been found so far in graves in Macedonia and in the north-northwest Peloponnese. The ninth leaf may have been written on in ink, now lost, as Pavlos Chryssostomou (1992) has proposed (G1). What emerges from the excavators’ reports is that, except for the unambiguous ivy leaves from Pelinna (D2) and for Philemena’s myrtle leaf from Elis which is also employed as a danake (F7), [75] the shape of the remaining seven leaves is either unknown or described as laurel or almond-shaped. In his study of the shapes of the incised leaves that have been published, Matthew Dickie has argued convincingly that the literary sources and the archaeological record allow for only a few trees to be represented by these gold leaves: primarily myrtle and ivy, and less often olive. [76] All three are evergreen trees associated with fertility, vegetation, and the chthonian aspects of Dionysos, Demeter and Persephone, Aphrodite, and also perhaps Athena. As Dickie discovered, and as Davaras, Despoini, and Kaninia also stressed in their studies of gold wreaths, [77] there is an insurmountable difficulty in distinguishing especially between myrtle-, laurel- and sometimes olive-leaves. This may be due to the shape of these leaves, all of them being oblong with differences in details which the ancient goldsmiths could not or did not care to reproduce. The goldsmiths may have been interested simply in a more schematic representation, letting the customer decide, or as Dickie aptly put it, letting the “context determine which plant was imagined to be represented.” [78]

There is, however, further evidence corroborating Dickie’s interpretation of these as myrtle-leaves, instead of laurel- or almond-shaped leaves, when the other grave-goods are taken into consideration. Together with the Poseidippos leaf (E3) [79] were recovered clay gilt myrtle-berries from a wreath and forty-one clay gilt pebbles in the shape of acorns, forty-six bone astragaloi, and a clay figurine of a female. The leaf of Philoxena (F6) was accompanied by clay myrtle-berries from a wreath and bronze gilt leaves, whereas Philon’s leaf (F5) was found with twelve lance-shaped leaves apparently from a gold wreath. The ivy-leaves from Pelinna (D2) were discovered with: a danake with gorgon in the deceased female’s mouth; a coin of Antigonos Gonatas; a diadem-like wreath of lead stem; gilt clay berries and gilt bronze myrtle-leaves with a gold ornament in the cranium; nearby a clay aryter; a clay bowl and two gold spirals ending in snake-heads; near the feet another clay aryter with a lamp inside it, a clay unguentarium, two bowls and a shallow skyphos; by the feet a bronze lebes with the bones of a neonate, probably the case of a baby and its mother having died in childbirth; and on the cover slab of the marble sarcophagus two clay bowls and fragments of a third, a clay feeder, and a clay figurine of a comic actor sitting on an altar. [80] The grave-goods recovered together with the unincised olive(?)-leaf (G1) are two gold myrtle-wreaths together with the bones wrapped in purple gold cloth inside the larnax which was placed inside a marble sarcophagus.

Statistically, these graves are few, but it seems that together with the gold leaf, the deceased was also ‘crowned’ by myrtle- and/or oak-wreaths (E3, F6, G1), either of pure gold or gilt clay. Given that laurel-wreaths in the archaeological record of Macedonia are scarce, because, as Despoini observed, laurel was the sacred tree of Apollo, whose relations with the dead and the Underworld were virtually non-existent, [81] and given that, in two cases, myrtle-wreaths were found together with the incised gold-leaves (E3, F6), it is rather unlikely that these leaves were meant to represent laurel-leaves. This is most probably also the case with the leaf from Aigeion (F5), although the possibility of the existence of different kinds of leaves cannot be ruled out, as the discovery of the ‘olive’-leaf indicates (G1).

A more or less similar picture emerges from the archaeological context of the remaining sixteen lamellae which have brief texts (Tables 1–2) and are either rectangular, ‘leaf’-, or mouth-shaped. Three graves with lamellae engraved with brief texts stand out specifically because of their grave-goods. In the grave of Phylomaga (no. 15 above; F3) the accompanying goods were: ivory fragments of the bier’s decoration representing floral patterns and figures of the Dionysiac cycle, [82] two gold finger-rings, and a bronze gilt wreath. Bottakos (F10) was buried with a bronze gilt wreath with berries; outside the grave to the northeast a trapezoid construction was found, apparently for the funeral supper (a similar case in D4, E5), and to the northwest of this construction were unearthed pottery fragments, traces of enagismos in later times, and a bronze gilt wreath with gilt clay berries. Inside Euxena’s grave (F1), [83] fragments of gold foils from a diadem were found.

Finally, of the nineteen remaining lamellae with long texts (Tables 1–2), four stand out in terms of their burial context. [84] Three burials at Thourioi (A1–3) yielded no offerings, except for the small hollows in the four corners of the chamber of grave 1, filled with ashes of bones and plants, indicating funeral sacrifices. The remaining two, however, are unique (A4, C1), in that the tumulus above the grave was nothing more than deposits of eight strata, each consisting of ashes, carbon and burnt pottery sherds topped by earth above, a strong indication of rituals, sacrifices, and hero-worship of the dead buried inside. Outside the grave only a few small black vases were found, and inside the chamber, where the cremation took place and the remains were simply covered by a white sheet which disintegrated when touched by the excavators, were found bronze locks of the coffin, two silver medallions on the chest decorated with ‘female heads’ (reminiscent of the head of Persephone to be found in images on the Apulian vases), a few small pieces of gold from the dress’ decoration, and two small wooden boxes with inlaid palmettes. The two gold lamellae were discovered one inside the other, A4 folded nine times and placed inside C1, which was folded like an envelope. The Petelia lamella (B1), according to Zuntz, was somehow recovered from its grave at a later period, rolled up, and clipped off, in order to fit in a gold case with chain attached, dated to second–third century CE; the final product was an amulet used for its “outstanding magical virtue.” [85] Lastly, the lamella from Pharsalos (B2) was discovered with two other small objects inside a hydria-urn which in turn had been placed inside a round limestone container. The urn, manufactured exclusively for burial use, is decorated at the base with ivy-leaves and anthemia, and below the neck-handle there is a representation of the ‘abduction’ of Oreithyia by Boreas, a scene essentially similar, according to Verdelis, to the abduction of Persephone by Plouton. [86]

This brief survey of the graves and their respective goods leads to some important points concerning the overall burial context of each grave. [87] In Macedonia, where more than half of the lamellae and epistomia discovered are incised with only a few words and where no lamellae and epistomia have been found bearing a longer text, shape did not matter, or so it appears. The epistomia of Hegesiska, Poseidippos, Philoxena, and the blank leaf (F11, E3, F6, and G1) are myrtle- or olive-leaves, as (most probably) are the ones from the Peloponnese (F2, F4, F5, F7). The shape of the remaining lamellae with short texts is either rectangular or rhomboid. There are also three extraordinary cases in Pieria, two gold coins employed instead of lamellae in Pydna (F8, F9), and another in the shape of a coin (a pseudo-coin?) in Dion (F12).

In certain cases, the deceased was crowned with wreaths of gold or of gilt clay, which were less expensive (E3, F5, F6, F10, G1), [88] or, in two instances, with a diadem (D2 and F1). In seven cases, the deceased appear to have received offerings, sacrifices, and enagismoi in some form of ritual after burial, with Timpone Grande standing out perhaps as a case of a local hero-cult (A1–4, C1, D4, E5, F10). [89] There are also four cases in which Dionysiac overtones emerge, irrespective of the text incised on the gold lamellae: D2 was accompanied by a clay figurine of a comic actor seated-on-altar in addition to the two incised ivy-leaves placed over each breast of the buried female; in the Timpone Grande at Thourioi, two medallions with a female head looking like the Persephone on the Apulian vases were placed on the chest of the deceased (A4, C1); [90] the Pharsalos hydria bears a representation of the ‘abduction’ of Oreithyia by Boreas, which reminds the excavator of the more familiar one of Persephone by Plouton and which “prepares and complements the text on the lamella” (B2); [91] finally, ivory fragments from the bier’s decoration in Methone represent figures from the Dionysiac cycle (F3).

The presence of gold or gilt clay wreaths inside a grave in addition to an engraved epistomion seems to defy explanation. Despoini has noticed that relatively few wreaths have been found in the hundreds of graves in Attica, despite the frequent references in literature and in inscriptions to honors individuals received, sometimes including a gold wreath, and other times including a myrtle crown after initiation at Eleusis. [92] Most of these, as with the gold athletic wreaths, [93] would have been dedicated to the appropriate god after the celebration and some would have ended up in graves. The number of wreaths, however, recovered from graves in Macedonia, indicates, according to Despoini, that the crowned deceased, reenacting the persona of either ‘an athlete’ or ‘a symposiast’ or (less likely in Macedonia) ‘a honored citizen,’ would have certainly expected to attain eternal life among the blessed. [94] The metaphor of the foot-race and the crowning of the victorious ‘athlete’ is employed in lines 9–10 of the lamella from Thourioi (A1): ἱμερτο̑ δ’ ἐπέβαν στεφά|νο ποσὶ καρπαλίμοισι. According to Zuntz, [95] the line is spurious because it is repeated in lines 12–14 of the same lamella, even though he admits that stephanos, the normal prize of victors, is appropriate for the occasion, and that it is used elsewhere, again metaphorically, to denote purpose and distinction. Bernabé and Jiménez San Cristóbal have discussed in more detail the literary references to events for which a wreath was employed and have concluded that the wreath-metaphor, rich in symbolism, may at the same time stand for mystic initiation, athletic triumph, and symposium. [96] But what about the deceased (especially in Macedonia) in whose graves both an incised epistomion and a wreath have been recovered?  [97] Is it sheer coincidence, or is this stephanosis nothing more than the material sign of the lines in the Thourioi text A1? Does this connote not only a metaphorical but also an actual event of the initiation? Would this burial practice in Macedonia thus allow us to associate the texts of group A with the ones in groups B and E? Whatever the case, the coincidence is indeed remarkable.

Likewise, the overall picture of “Men and Women, Rich and Poor” (to borrow Graf’s title  [98] ) that emerges from the graves where lamellae and epistomia have been found is not clear, as the evidence is inconclusive and perhaps misleading. In terms of gender, males and females are equally equipped with the necessities for their eternal trip. There is no evidence to suggest any preferences, besides the usual, of males or females regarding certain types of offerings or concerning the texts incised. In terms of affluence, richer graves may give an impression of a ‘Dionysiac context,’ while poorer graves may allude to an ‘Orphic/Pythagorean’ one. The archaeological finds in the graves at Thourioi (A1–4, C1) are moderate if not austere in comparison to the finds in some graves at Macedonia, but grave-goods at Thourioi may have been of less importance as compared to the tomb’s actual construction and the later ritual over the tumuli. Even so, as Themelis and Touratsoglou have shown convincingly with regard to the Derveni grave, generalizations and prima facie conclusions should be resisted. The discovery of the papyrus in Derveni grave A led to the conclusion that the deceased was an Orphic follower. The richness, however, and the strong Dionysiac character of the grave’s archaeological context stand in sharp contrast to the Orphic-Pythagorean austere life and to its more moderate means. [99] As has become evident only recently, ‘Orphism’ and Dionysos are not, after all, mutually exclusive. The graves with a lamella at Hipponion, Thessaly, and Sfakaki constitute an intermediate stage between the ‘rich’ burial customs in Macedonia and the ‘poor’ ones in Thourioi. [100] The grave-goods recovered from the five graves with an epistomion in the Sfakaki cemetery indicate that the deceased were of moderate means; these five graves are not among the richest of the 56 excavated at Sfakaki so far, but they are richer than the pit-graves. The overall picture of the undisturbed graves is that of a careful and well-ordered burial but with no extravagance.

A key issue related to the lamellae’s shape and apparently a matter of importance to those buried with an epistomion was the placement of the lamella inside the grave. Some are found near the cranium or the mouth (A4, C1, E4, F7, F8, F9), others on the chest (B10 [101] , D2, D3?, D4), or, less often, close to hand (A1–3, F3). What is striking is that the so-called Charon’s obol or danake was placed inside the mouth, at least from the second half of the fifth century BCE onwards, [102] but it was also placed in the hand, on the chest, or simply anywhere inside the grave, a practice that Guarducci associated with the incised lamellae. [103]

Not all graves, however, contained a coin, and not all graves with an epistomion contained coins—these two facts imply a differentiation in burial practices and funerary ideology. Placing a coin in a grave is not a widespread phenomenon within the ancient Greek necropoleis. This practice should therefore not be associated exclusively with Charon’s Greek myth, because it does not fit entirely well with this myth, and because it is also attested in other cultures where the Charon myth does not exist. The first evidence of this practice, so it appears, comes from the famous scenes of Aristophanes’ Frogs between Dionysos, Xanthias, and Heracles (lines 139–140), and between Dionysos, Xanthias, and Charon (lines 170–270). The ferryman transports the dead for a fee to the Underworld, where, instead of the fee, Dionysos is forced to pay in kind, ‘working’ as an oarsman, a theme that will be later developed and expanded. There is no doubt that this was an actual practice at the time of Aristophanes—otherwise the scene’s jokes and hilarity would be pointless—but this does not confirm how widespread it was, nor does it answer why only a rather small group of people practiced it. Although she does not discuss the issue of the burial-coin practice, Christiane Sourvinou-Inwood has argued convincingly that Charon the ferryman and Hermes Chthonios emerge as psychopompoi in art already by ca. 500 BCE. [104] This indicates shifting attitudes and ideologies in the archaic period as new needs arose, either from a development of a more individualistic attitude towards and concern for death, the afterlife, and its rituals, or from the emergence of a polis system which looked to control burial practices as well as funerary rituals and ideology.

Keld Grinder-Hansen (1991) proposed to replace expressions like “Charon’s/Charonian obol/fee/coin” with the less ideologically-charged “death/burial-coin” or the like, whereas Susan Stevens rightly stressed that references or allusions to “Charon’s obol” in a variety of texts are guided by different aims, all of which imply a connection between poverty and death, as the obol is the cheapest denominator. Thus, this expression is employed for humor or an ironic look at the vanity of conventional views on the afterlife, but it also signals the replacement of alimentary goods in the grave in exchange for the nourishment of the soul as it begins its journey. When the coin is placed inside the mouth immediately after death, it may especially denote “a rite of passage rather than burial practice.” [105] This interpretative variety is also exhibited in the archaeological record and, according to Stevens, it comes from a belief rooted in the religious-magical significance and intrinsic value of coins on account of their ‘invisible’ power. This burial practice or rite of passage was “a way for the living to communicate with the dead, to promote life among the dead, while the door to the other world was still open.” [106] Renata Cantilena has correctly remarked that change in terminology provides a more accurate description of the facts, but does not solve the essential problem of explaining the funerary ideology, if any, behind this burial practice. [107] Placing a coin in a grave has indeed been explained in many different ways: it may or may not indicate the affluence of the deceased and his or her social status as another burial offering; it may constitute a symbolic payment or recompense facilitating the passage from life to death; it may also have been used as a talisman to protect the dead or as an amulet for protection of the living against the dead; or even, as Rhode had proposed, [108] as a pars pro toto, symbolizing the transference of the dead’s wealth to the living members of his family. [109]

These explanations, alongside with others that account for economic, political, and social circumstances, need not account for every coin in every grave. They simply bring to the fore some of the ideas and symbolism that people may or may not have had in relation to the burial-coin practice. Sourvinou-Inwood’s recommendation of “more complex and ambivalent categories” to replace a “dichotomy belief/not belief of the Greeks in the myth of Charon” which is “culturally determined and misleading” [110] is applicable mutatis mutandis to the practice of the burial-coin, and arguably to the use of the gold lamellae and epistomia as well. In particular, there are cases (D2, D4, E4, F2, F4, F5, F7, F8, F9, F12?, G1) where both a lamella and a coin or pseudo-coin accompany the deceased, but there are cases where only a lamella is found, and when it functions also as an epistomion, the lamella apparently takes over the coin’s duties altogether.

For these difficult issues, the excavated part of the Sfakaki cemetery may serve as a test case for a rule of thumb. Of the 56 burials excavated so far, five graves contained gold lamellae (nos. 8–12 above; B12, E4, G2–4). All of them, regardless of shape (‘mouth,’ rectangular, rhomboid), were epistomia, i.e. they were found in the mouth or near the cranium, and in one case, a bronze coin was found on the chest of the deceased (no. 8 above; E4). One is dated between the third and early first centuries BCE and four are dated to the first century CE, and they comprise three groups of gold lamellae, just as the epistomia from Macedonia, Thessaly, and South Italy (Tables 1–2): unincised (group G), incised with a few words (group E), and incised with a long text (group B).

The epistomia from Sfakaki together with their grave-goods demonstrate similarities in burial customs—whether this might also indicate familial relations of the deceased must remain a conjecture—but at the same time, they militate against generalizations. Of the twenty-six graves studied so far, including those with epistomia nos. 8, and 10–12 above, twenty–two contained burial-coins, but of the four graves with epistomia , in only one grave was a coin discovered on the chest of the young male (no. 8 above; E4). This may be accidental, but it may also be that the deceased with epistomion no. 8 and his relatives felt strongly about the burial-coin practice, whereas the deceased with nos. 10–12 (G2–4) probably employed these three unincised epistomia as pseudo-burial-coins, because of their intrinsic value, and at the same time perhaps as unincised tokens of initiates for passage and transfer to a special place of the Underworld.

Margarita Guarducci was the first to realize the similarities between the custom of placing a coin in the mouth and the customs seemingly surrounding some of the epistomia. [111] Accordingly, she postulated a practical explanation. On account of its shape and the fact that it is not folded, lamella no. 4 above (B6) was probably placed at the right hand, as the Thourioi lamellae (A1–3), whereas lamellae nos. 1–3 above (B3–5), on account of their being folded, were probably placed inside the mouth, the safest place of the body. Guarducci’s suggestion encountered Zuntz’s scepticism, [112] because epistomia nos. 1–3 above (B3–5) were not unearthed during systematic excavations. Puzzled, however, by the fact that some lamellae were found folded or rolled up, so as to ‘become coins’ and fit into the mouth in order to ‘put the right words on the tongue,’ Zuntz allowed for the possibility that some of these may have been later employed as amulets. There is, however, no substantial evidence whatsoever that these were put inside cases, except for the curious case of the Petelia lamella (B1), and for the reports of the Eleutherna sellers for nos. 5–6 above (B7–8). [113] In like manner, Petros Themelis (1994) suggested for the gold myrtle-leaf incised with the female deceased’s name (F7) that, since no coin was found inside the grave and the leaf was discovered under the cranium, the incised myrtle-leaf may have also served as a danake.

Guarducci’s and Themelis’ cautious suggestions are corroborated by three unique (thus far) examples in Macedonia, which bring together the use of the burial-coin practice with the mystic symbola. Matthaios Bessios (nos. 13–14 above; F8, F9; Figures 14–15) reports the discovery of two gold coins of Philip II incised with a male and a female name: Andron and Xenariste, respectively; the coins were found in the mouths of the deceased, buried in two almost identically decorated graves. [114] Dimitrios Pantermalis has published from Macedonian Grave V in Dion a photograph of a small gold disc on which the name Epigenes was incised with dotted letter-strokes (F12). [115] Coins on which personal names are incised are extremely rare: either because of a lack of a gold foil, or lack of time, or for some other reason, the relatives(?) of Andron, Xenariste, and perhaps Epigenes employed two gold coins and a small gold disc (as a token or as a pseudo-coin?) on which they engraved the names. These three examples appear to combine (in a manner so far unique) the burial-coin with the gold lamella practices. [116] A comparable but not entirely similar case is presented by a gold rectangular tablet with an inscription addressing Serapis, found inside a skull in a cinerary urn in Columbarium III at Rome. Although it is not Orphic, this phylactery, with its address to Serapis, presents (according to David Jordan) a curious case of either a Charonian obol or a mystic symbolon. [117]

The ambiguity between a burial-coin practice and a mystic symbola lingers, and the Sfakaki epistomia, two of them incised with different texts and three unincised, only increase the difficulty in approaching the problem. In spite of all the information they yield, evidence remains inadequate and a number of practical problems persist. Answers to the questions below would seem crucial in accounting for the differentiation; at the moment, unfortunately, we can only answer them with conjectures:

The nature of the evidence surrounding the five epistomia from the graves at Sfakaki allows only for assumptions and educated guesses in these matters. In spite of the graves’ similarities in terms of place, date, and content, no conclusive and convincing explanations may be offered regarding burial customs and rituals, and their funerary ideology. The five epistomia from Sfakaki attest that differentiation in burial customs may have been both a diachronic and a synchronic phenomenon, and they seem to embrace diversity or individualization, as Sourvinou-Inwood (1995) has argued, rather than homogeneity. Among the five graves at Sfakaki comprising a deceased buried with an epistomion, three different practices are evident. This variety suggests that, even though these people were inhabiting the same area, practiced similar burial customs, became mystai in a Bacchic-Orphic cult promising life after death, and lived one sometime between the third and the early first century BCE (B12), another between 25 BCE and 40 CE (E4), and three in the first century CE (G2–4), they nevertheless developed a more personal attitude toward death.