At the time of its re-founding on the slopes of Mt.Thorax in the early 4th century BC, the city of Magnesia would have encountered a rural landscape that was already ripe with meaning. M. Nilsson emphasized that in the popular imagination, the gods were already present in nature, before planners and architects superimposed their rationalized vision of urban life:

Anyone who wishes to understand the religion of antiquity should have before him a living picture of the ancient landscape as it is represented … in Strabo’s description of the lowland at the mouth of the river Alpheus. “The whole tract,” Strabo [(VIII.3.86)] says, “is full of shrines to Artemis, Aphrodite, and the nymphs, in flowery groves, due mainly to the abundance of water; there are numerous hermae on the roads and shrines of Poseidon on the headlands by the sea.” One could hardly have taken a step out of doors without meeting a little shrine, a sacred enclosure, an image, a sacred stone, or a sacred tree. Nymphs lived in every cave and fountain. This was the most persistent, though not the highest, form of Greek religion. (Nilsson 1961, pp.17-18)

Such practices illustrate the concept of “lived religion” as applied to the ancient world (McGuire 2008; Raja and Rüpke, 2015, pp. 3-4). The formal sanctuaries of institutionalized religion were but one aspect of the space of ritual in cities like Magnesia. The complex polytheistic underpinnings of the region were not completely subsumed within the axial grid that was imposed when the city was founded in the Classical period. In fact, these older relationships – between diverse gods, their shrines, and the landscape – may help us to form a more complete understanding of Magnesia’s urban layout. Because archaeological remains of the city plan are scant, the topographical data contained in inscriptions describing cultic practices in the vicinity provide an alternative means of mapping the city. M. Detienne (2002, 148) argues that this ‘grass roots’ approach, starting from “the objects, the acts, the particular situations presented by the primary data, using these as so many reagents to see what aspect of this particular divine power comes to the surface in a given configuration” can help avoid the trap of over-generalizing the symbolic interpretations that the presence of a temple devoted to a given deity might evoke. Allowing “an experimental dimension somewhat analogous to qualitative analysis in chemistry” to guide such investigations leads to the identification of “a range of possibilities, without which polytheism remains opaque, a dead system (ibid.ibid.).” It is precisely such a range of possibilities for the spatial and architectural configuration of Magnesia that the present work seeks to elucidate from the textual, archaeological, and topographical data.

I will focus my discussion on the evidence concerning the lesser-known rituals surrounding Apollo of the cave at Hylai and Dionysos “of the plane tree”. These two cults, particularly when considered in conjunction with one another, will enrich the experimental reconstruction of Magnesia’s urban topography which is the purpose of this study. Movement is a critical factor here, located through multi-modal analysis of the rituals that connected the two religious centers, that brings the reconstruction into the third (spatial) and fourth (temporal) dimensions. Significantly, this was movement that crossed the boundaries of the city walls, constituting a multivalent perspective that encompassed the city and its architecture both from within and without. This chapter works from the precepts set out by Nilsson and Detienne to uncover the overlapping polytheistic dynamics that are reflected in Magnesia’s urban development.

The Cave of Apollo at Hylai

From the earliest time of its formation, Magnesia was associated with a cave sanctuary dedicated to Apollo. It is possible to reconstruct the outlines of this cult from several pieces of textual and numismatic evidence. The first indication we have for the cave-dwelling of Apollo comes from the Letter to Gadatas written by the Persian king Darius sometime during his reign (522-586 BC) in which the latter reprimands Gadatas for exacting tribute from the “sacred gardeners of Apollo”:

The king of kings, Darius, son of Hystaspes, to his servant Gadatas speaks as follows: I understand that you do not obey every point of my instructions…. Since you choose to disregard my desires as regards the gods, I shall cause you to experience, if you do not change, my wrath excited by an injury. The sacred gardeners of Apollo have been subjected by you to tribute and required to work profane land; that is to disregard the sentiments of my ancestors toward the god who said to the Persians …… [1]

The letter is known from a second-century AD Greek inscription, now located in the Louvre, found on a marble stele in villager’s garden in present-day Germencik, which lies on the road between Magnesia and Aydin (ancient Tralles) about 6 km east of the city of Magnesia. The text is generally accepted as a second century A.D. Greek “re-publication” of an actual Persian-era letter, although some (Hansen, 1896) doubt its authenticity due to the apparent irony in Darius’ professed reverence for Apollo – during the reign of Cyrus, the Persians had burnt the great temple of Apollo at Didyma to the ground. However, others (Briant, 2002) argue that the tone of the letter is consistent with Persian epistolary style from the period. Whatever Darius’ motivations for protecting the gardeners of Apollo, the letter seems to have been re-aired as a confirmation of the special privileges held by them, which would have been further reinforced by the great antiquity of their sanctuary and its status. At the time of Darius, Magnesia would have been situated at its original location somewhere in the Maeander valley. We can therefore draw the conclusion that a cult of Apollo in the vicinity of the current site pre-dated the city itself. After the new city had been established, evidence for a cult of Apollo involving dendrophoroidendrophoroi – men who carry trees – appears on Roman coins of Magnesia of the 3rd century AD (Robert, 1977, p.78) (Fig.13). The cult is attested as late as the end of the 5th century AD by a reference to Appollonos Aulai in the life of Isidore by Damascius. If all of these documents do in fact refer to the same cave cult, then it would have existed for at least 1,000 years in the vicinity of Magnesia.

Images of dendrophoroi from the coins of Magnesia. Left: Gordian III (225 – 244 AD); Right: Otacilia (244 – 249 AD). After Robert (1977, Fig.15, p.78).

The location of the cave, however, remains a mystery. Texier (1849, 90) was confident it could be discovered in the hills to the north of Magnesia, near the village of Gümüş:

Following the slope of the mountain towards the village of Gumuch, halfway up one can see a large cave open to the south. I have not found in the these ruins any inscription that could tell me the ancient name of the village; but this cave is, I think, sufficient evidence to recognize the village of Hylae, which, according to Pausanias, was Magnesia’s neighbor, and distinguished by a cave dedicated to Apollo in which it preserved a very ancient statue of the god [2].

Rayet (1877, 132) confirmed that Texier’s conjecture was “probably accurate”, while conceding that “I haven’t seen it myself, and I haven’t learned that there existed in the environs of Magnesia any other remarkable cave”. Kern (1895, 93 n.1) dismissed this conjecture, asserting that while there are actually three cavities above the village of Gümüş, “those caves are nothing more than quarries”. Moreover, he had “searched for this cave very often, but always in vain”.

The toponym for the site of Apollo’s cave was identified as Hylai in Pausanias; an alternate transliteration was proposed by Wilamowitz (1900, p.572, n.3) which changed the name to Aulai in order to bring it in correspondence with Magnesian coins of the 1st-3rd c AD which bear the epithet “Apollo Aulaites” and depict the figure of Apollo Chitharoedus (Schulz 1975, p.39). Robert (1977, p.83) supports Wilamowitz’ revision, and ‘Aulai’ has since been widely adopted in translations of Pausanias. However, Ustinova (2009, p.120) points out that Hylai could derive from Hyllouala in Caria, where there was a sanctuary and oracle of Apollo. Whether or not these were one and the same, the indigenous Carian etymology of Hyllouala would support the conjecture that the cave-dwelling Apollo was a Hellenized incarnation of an older god. Ustinova assumes that the sanctuary would have functioned, like many dedicated to Apollo, as a manteon or oracular site. While there is no evidence in the texts that suggests the “gardeners of Apollo” did anything except enact a form of oribasie, the exalted frenzy that led them to uproot and carry whole trees, it should be noted that most natural or architectural caves dedicated to Apollo did have some oracular function (Ustinova, 2009, pp.109-155). Ustinova further proposes that Hylai could also be identified by a place called Hieracome, mentioned by Livy in the time of Augustus (38.13):

[From Magnesia] after crossing the Maeander they marched to Hiera Come. Here there was a noble temple to Apollo and an oracular shrine; it is said that the priests delivered the responses in smooth and graceful verses.

Apart from being mentioned after Magnesia, there is nothing in this passage which corresponds with our picture of Apollo and the dendrophoroi, nor does Hieracome bear much etymological resemblance to Hylai. By all other accounts, Apollo lived in a small cave, and there is no mention in Pausanias of a “noble temple”, or an oracle. Furthermore, the behavior of the dendrophoroi doesn’t harmonize with the delivery of “smooth and graceful verses”. The geographical detail of Hieracome being reached after crossing the Maeander also doesn’t agree with the other evidence. If Apollo’s cave was in the hills near the south bank of the Maeander, it would be at a distance of at least 8 km from the city, rather far for the dendrophoroi to carry their trees (not to mention necessitating a river crossing as well) [3], whereas their presence on the coins of Magnesia suggests that they were associated with the city’s identity and probably geographically proximate as well. Furthermore, in the last known account of the sanctuary at Aulai/Hylai [4], in the life of Isidore by Damascius, two late-5th century philosophers make a pilgrimage to the sanctuary from their home in Aphrodisias. The narrative describes them swimming across the Maeander to reach the sanctuary and almost drowning [5]. Since Aphrodisias is located south of the Maeander, this would seem to indicate that the Hylai was north of the river, in other words where the city of Magnesia was located.

Despite the difficulties in pinpointing the exact location, there are certain conclusions that can be drawn with confidence. One is that a sanctuary dedicated to Apollo existed in a cave outside the city in the hills or mountains, not only because this is geographically where caves are found but also because of the “sheer precipices” and “narrow paths” described by Pausanias. It is clear that the cave was in a wooded, rural area that was nevertheless proximate to the city of Magnesia, since the dendrophoroi were uprooting fully-grown trees, yet are depicted on many of the coins of the city, thus placing them firmly within the imagery associated with the city’s identity. It is likely that Hylai was located on the slopes of Mount Thorax instead of the hills to the north (above the village of Gümüş) because of the higher and wilder elevations of Thorax. There is archaeological evidence in support of this theory as well. A statue of Apollo, clad in the long Chitharoedus garment in which he is depicted on the coins bearing the epithet Apollonios Aulaites, was found in the hills southwest of Magnesia, near Argavlı village, in 1995 (Bingöl 2007, 179). The proximity of the location of the find with the nearby “Büyük Manastır” site, where Thibron is presumed to have brought the Magnesians to safety before relocating the city to the plain, is further evidence in support of the conjecture that the fortified hilltop site was known as Hylai [6].

The great antiquity of the sanctuary, established by the “Letter to Gadatas”, supports the theory that the name was derived from an indigenous Carian toponym. The site of the cult and its members enjoyed special privileges that were seemingly preserved until late antiquity, as indicated by the extracts from the Life of Isidore. The pre-Hellenic roots of the cult, perhaps, account for the eccentric behavior of the dendrophoroi and their deviation from the normal settings and practices of a sanctuary dedicated to Apollo, in which the priests did not practice divination, but instead engaged in rites more commonly associated with Cybele or Dionysos. Because the city was surrounded by defensive walls since Hellenistic times, the dendrophoroi would have to enter the city via one of the gates, thus connecting their ritual with the city layout and street grid. The specific nature of this alignment will be closely examined below, using the 3D model to investigate the ways in which the cave of Apollo might be traced in the urban scheme of Magnesia through its possible connection with a group of sanctuaries dedicated to Dionysos.

Cult of Dionysos

Like Apollo at Hylai, we find Dionysos behaving rather out-of-character in Magnesia. Although not explicitly connected with Apollo Aulaites in the epigraphical record, the cult of Dionysos occupies an intriguingly complementary position when considered in relationship to the urban topography. An inscription found in Magnesia in 1890 records a singular event that would have occurred in the early 3rd century BC (Kern, 1900, no.215) [7]: One day a large plane tree standing in the city center was torn apart by a violent wind, revealing an image of the young Dionysus. Upon consulting the oracle of Apollo at Delphi as to the meaning of this occurrence, the Magnesians were told to bring back with them three Theban maenads who would instruct them in the ways of Dionysus and his rites, for they had forgotten the god, who was already present when they built their city.

This inscription comprises two marble pieces: a larger stele that recounts the oracle, and a smaller square piece containing a dedicatory inscription. The two pieces were not found together, being scattered in various parts of the village of Tekke, but Humann and Kern assert that they have it on good evidence that both pieces were originally found at a site just to the west of the City Gymnasium “at a point on the field where the debris clearly indicates that an antique building formerly stood there” (Kern, 1895, p.85). This location, now completely covered, is also the site where in Humann and Kern’s time another inscription could still be observed in situ, and which they called the “Mysteninschrift” (Kern, 1900, no.117). This inscription testifies to the practice of Dionysian mysteries and records the sums of money donated by worshipers. On the basis of these three inscriptions, Kern and Humann posit the existence of a temple of Dionysos at this location, a few steps west of the City Gymnasium. This location, in the heart of the city, would accord with the account of inscription no.215, which describes the plane tree where the epiphany of Dionysos happened as πλατάνου κατὰ τὴν πόλιν. The inscription contains a prose postscript (Kern, 1900, no.215a) which reads:

"In accordance with the oracle, and through the agency of the envoys, three maenads were brought from Thebes: Kosko, Baubo, and Thettale. And Kosko organized the thiasus named after the plane tree, Baubo the thiasus outside the city, and Thettale the thiasus named after Kataibates. After their death they were buried by the Magnesians, and Kosko lies buried in the area called Hillock of Kosko, Baubo in the area called Tabarnis, and Thettale near the theater [8]."

The genuineness and dating of the inscription is crucial if we are to take it as a guide to the actual topography of Magnesia. Though controversial, the oracular origin of the Dionysian cult is attested by multiple chronological details, and thus the inscription provides valuable information on the urban layout of Magnesia. Henrichs (1978, p.130) provides a convincing analysis of the text and concludes that “we may now proceed on the assumption that the city of Magnesia, at the urging of the Delphic oracle, actually imported three maenads from Thebes sometime between 278 and c.250 BC, and that these maenads died in Magnesia and were buried there at public expense, probably before 207/06 B.C.”

The inscription names the three thiasoi (groups of devotees) led by the Theban maenads, along with their burial places, but finding the places they refer to is no simple matter [9]. The maenad called Kosko led a thiasos called Platanistinoi, which seems to be a reference to the plane tree where the epiphany of Dionysos occurred. If, following Humann and Kern, we accept that the supposed Temple of Dionysos was constructed on the site where the plane tree originally stood, just to the west of the Gymnasium, this may be the location of Kosko’s thiasos. She is described as being buried at Koskobounos, which suggests it was on a hill. This toponym raises the question of whether the maenads’ burial places are to be sought near the location of their respective thiasoi, since the terrain around the putative temple of Dionysos is relatively flat. On the other hand, the plane trees invoked by the name of Kosko’s thiasos need not have been on the site of the epiphany, but may have simply referred to a grove of plane trees elsewhere where the group of devotees gathered to pay homage to the original appearance of Dionysos in Magnesia.

Furthermore, the name of the second thiasos places it in contradistinction to the city and therefore may indicate that the other thiasoi were indeed located, by contrast, within the urban environment. The second maenad, Baubo, led a thiasos pro poleos (outside the city walls) and was buried at a place called Tabarnis. Henrichs (1978, 134) suggests that Baubo’s thiasos might have been the most likely of the three to host actual maenadic rituals, since it was located at a proper distance from the city. Dionysiac rites were traditionally associated with natural scenery such as woods and hills. The site associated with the toponym Tabarnis remains unlocated, although Kern assumes that Tabarnis must be “a place outside the city” (Kern, 1895). In fact, Tabarnis was mentioned in another Magnesian inscription, which describes it as being the location of a spring from which water was diverted to the city (Kern 1900, no. 251). This suggests at least two possible locations: an aqueduct descending from the hills met the city walls in the south-west (Humann, 1904, p.29). This interpretation would seem to indicate that Tabarnis was a village within Magnesia territory, situated in the foothills of Mt.Thorax. Alternatively, Humann’s plan locates a spring above the Theatron, directly in line with a pump room and fountain in the southwest corner of the Agora that was excavated by the German team. This spring is within the city walls, but would have been outside the city grid proper, as the streets and residential blocks probably did not extend into the foothills (Bingöl 2007, p.128). On the other hand, Reinach (1890) argues that the words pro poleos (which he interprets as “before the city”) must be taken to indicate a location of a sanctuary at the gates of the city, while Henrichs (1978, p.130) deems it “very likely” that Tabarnis echoes the Latin taberna, suggesting an Imperial, rather than Hellenistic, date for the toponym [10]. According to Henrichs, this does not rule out the actual existence of such a place, as the author of the inscriptions, Apollonios Mokolles, may have simply been describing the location of the ancient maenadic tombs using the name he was familiar with from his own era. If Tabarnis is indeed derived from taberna, this points to a location within the city where shops would be found, rather than outside of it. However, the combined evidence of the inscription that identifies Tabarnis with a spring that supplied the city with water, together with the designation of Baubo’s thiasos pro poleos, seems to indicate that Tabarnis would most likely be found in the southern foothills above the city.

The etymology of the toponym describing the location of the thiasos of the third maenad Thettale, kataivatai, is as suggestive as it is controversial. The early commentators were drawn to make the connection with katabasis and descent to the underworld: Reinach said that “the thiasos of katavatai awakens the idea of descent and particularly of the descent into hell”. Reinach evokes Strabo’s mention of several charonion, or caves which functioned as portals to the underworld, in the vicinity of the Maeander valley (Strabo 12.8.17). After pointing out that the name of the Lethaus River evokes Lethe, one of the five rivers of Hades, he continues: “perhaps, however, katavatai means simply an area of Magnesia which sloped steeply towards the river Lethaios” (Reinach 1890, p.360). Kern (1895, p.93) also argued for a meaning of ‘descent’, citing the thasis ton Kataivaton, an underground crypt where the Eleusinian mysteries were held. However, instead of transposing this on the sloping banks of Lethaios, he prefers the steep mountainsides of Thorax. Again he makes the association with a cave, but this time rather than the charonion of Strabo he mentions the fabled cave of Apollo at Hylai, and reiterates this connection in the catalog of inscriptions published in 1900 (p.139, note to line 36). Albert Henrichs (1978, p.133) instead takes kataivatai as a reference to Zeus Kataibates, or Zeus who descends in lightning: “they would have met at a location where the lightning of Zeus had struck, and was therefore taboo except for religious use” [11]. He also makes the interesting point that kataivatai is a masculine form; suggesting that Thettale’s thiasos may have included men: an unorthodox configuration. It is tempting to speculate that the men in this third thiasos were the dendrophoroi of Apollo, and that katavatai is indeed a reference to descent into a cave, or alternatively, from the heights of Aulai.

Possible locations of the three thiasoi.

Connection between Apollo of the Cave and Dionysos in the City

By the late 19th century, Reinach had first noticed the connection between the Dionysian thiasos of katavaiton and the idea of “descent” (as into a cave), and Kern had explicitly linked this to the cave of Apollo at Hylai, The name dendrophoros had been used in numismatic catalogs to identify the men carrying trees pictures in numerous Magnesian coins (Schulz, 1975, p.39). However, this appellation was understood as a reference to the followers of Cybele, and not Apollo, since the cult of Cybele was known to use tree branches in their ceremonies. A direct link between Apollo and Dionysos at Magnesia, however, remained elusive. In 1895, numismatist Imhoof-Blumer published a Magnesian coin of the mid-2nd c. AD which shows a child Dionysos (a type of representation of the god frequently encountered at Magnesia) seated in a shrine between two columns, and figure dancing before him, carrying a tree (Schulz, 1975, p.76 no.189). The two columns seem to indicate the existence of a temple of Dionysos in Magnesia which housed the image of the god, just as Humann had suggested. Imhoof-Blumer identified the dancing figure as the dendrophoros, citing the passage in Pausanias that describes the men who “uprooting trees of exceeding height walk with their burdens down the narrowest of paths”, and links it with the inscription of Dionysos’ appearance in the plane tree. He proposed that “the who bring their burden to the sanctuary of Dionysos indicate a ritual in honor of Dionysos Dendrites, the god of vegetation” (Imhoof-Blumer, 1895, pp.285-286). In 1977, a comprehensive study of the connection between Apollo in the cave and Dionysos in the city was offered by Louis Robert. His review of the numismatic evidence for representations of Apollo and Dionysos in Magnesia (Robert 1977, p.84) led him to conclude:

It seems reasonable to me to assume that those possessed by Apollo, drawing their extraordinary powers from the statue of Apollo at Aulai, doubtless began their exploits of running through the mountains and uprooting of trees at the sacred wood of the god, and wandered uninterrupted with their charge through the mountains and ravines of the surrounding area (including Mount Thorax) to arrive in the city at the sanctuary of Dionysos and deposit before the god the uprooted tree. It was necessary that this tree, uprooted at Apollo’s cave in the forests of Aulai, arrived somewhere. It cannot be brought back to the sanctuary from whence it had been removed, since the “obstacle course” exhausted those who had brought it far away. One cannot abandon it in a corner.

He notes that the cults of both Dionysos and Apollo in Magnesia are united by their connection with trees: the plane tree in which the image of Dionysos was found, and the trees carried by the followers of Apollo. Since the cave at Aulai predated Dionysos’ epiphany, perhaps the ritual of the tree-carrying dendrophoroi was a response to this event, a re-enactment of the appearance of Dionysos in seasonal celebrations.

Certainly the specific connection between Aulai and Dionysos’ temple can only be as old as the temple itself. Before that, we must still account for the cave and its strange gardeners. Though not unprecedented, the cave is an unusual home for Apollo, who is more commonly encountered as the god of architecture and cities [12]. Apollo Pythios played a role in the founding of Magnesia through his oracle at Delphi. In fact, the dynamic between Apollo and Dionysos at Magnesia flouts most of the modern stereotypical characteristics of both gods, although these stereotypes (such as that Apollo represents reason, order, intellect, etc.) may be mostly modern inventions. Marcel Detienne (1986; 2002) sees the persistent Apollonian-Dionysian duality in modern thought as the result of the influence of Friedrich Nietzsche’s Birth of Tragedy. This work, Detienne argues, eclipsed the complex relationship that existed between Apollo and Dionysos in many local variations throughout the Greek world (Detienne 2002, p.148). He proposes instead that the Apollonian-Dionysian dynamic be treated as an example of a polytheistic system “primarily constituted by the relations between gods”, characterized by “practices and their variety…concrete configurations…, the acts, objects and situations that contextualize the relations between divinities”. Such a configuration may be sought in the spatial and architectural (or natural) context in which the ritual of the dendrophoroi was played out at Magnesia. It is by teasing out these concrete, empirical localities, architectonic settings, and movement between them, that we may begin to understand the relationship of Apollo and Dionysos as it existed at a specific place and time.

In view of this, the fact that Apollo is to be found in a cave at some distance from the city, is worth examining more closely. One strain of Apollo’s character is thought to have its origins as an indigenous god of Asia Minor, who originated in Lycia and was Hellenized by around 800 BC (Weinberg 1986, p.134). As such he originally was much more ‘unstable’ and chthonic in character; only after the progressive taming of his nature towards the qualities more familiar from the Nietzchiean perspective does he seem to represent the diagrammatic opposite of Dionysos (Weinberg, 1986, p.305). In fact, the distinction between Apollo and Dionysos is often blurred or even nonexistent, particularly in the sense in which both gods functioned as “gods of inspiration”; whether towards prophecy or frenzy or a combination or the two. At Delphi, the center of the Greek world (and a place which held a particular importance for the Magnesians, who frequently consulted the oracle there), the site is shared by Apollo and Dionysos. There is evidence that before Apollo and his cult arrived at Delphi, the place was sacred to Dionysos (Weinberg 1986, p.304); even after Apollo’s accession, Dionysos remained the sole occupant for the three months of the year that Apollo was said to be visiting the Hyperboreans in the north [13]. Significantly, the Delphic oracle may have drawn its power from a cave, although ancient accounts and modern archaeology differ as to whether this cave actually existed and what function it served in the sanctuary [14]. Strabo (9.3.5), although apparently without having witnessed it himself, recounts: “They say that the prophetic chamber is a cave, hollow in depth, with a rather narrow mouth, from which arises the breath of inspiration.” However, modern archaeologists have not found evidence for a geological formation corresponding to this description. Whatever the form of the cave, the existence of this configuration at Delphi, a place of manifest importance for the Magnesians, may be the template for the recurrence of the Apollo – Dionysos – cave triad in Ionia.

Chthonic and Underground Cults at Magnesia

Spatially, the cave is as elusive at Magnesia as its idea is evocative. Still, the prevalence of underground spaces associated with caves or cave-like functions in Greece allows us to entertain the possibility that some resonance of the cave of Apollo at Hylai might have found its way into the architecture of Magnesia. Caves dedicated to Apollo were relatively rare, though not without precedent. Delphi is the most famous example, and in Asia Minor the temple of Apollo at Hierapolis and the temple of Apollo at Claros both contain underground spaces beneath the cella where prophecy was performed. The capacity of caves to cause the phenomenon of enthousiasmos – inspiration through possession by the gods -- was well-established in antiquity. Strabo (14.1.1, 12.8.17) refers to three such caves in the Maeander Valley: one of which, the Aornum, was at a place called Thymbria halfway between Magnesia and Myus. This cave, along with that of Acharaca outside Nysa, and the cave under the temple of Apollo at Hierapolis, was one of a type that seems to have been filled with noxious vapors that produced mild hallucinations in those who entered. All three of these caves were considered Charonia, or gates to the underworld, and indeed the cave at Acharaca was associated with the temple of Pluto and Kore. However, the consciousness-altering properties of each cave served a different function: at Acharaca the cave had healing properties, while at Hierapolis the vapors may have served an oracular purpose (Ustinova 2002, 284; Ustinova, 2009, 273). At Aezani in Phrygia, the cave sanctuary was associated with an indigenous mother goddess who was Hellenized as Cybele.

Turkish archaeologists have recently discovered that the east stoa of Magnesia’s agora, which contains the propylon to the Artemision sanctuary, originally possessed two levels, the lower one being an underground cryptoporticus (Bingöl 2007, pp.105-109). The agora is heavily impacted by the high ground water that regularly floods the area, and it is hard to imagine why this underground structure would have been built if not for some important ritual purpose. This conjecture is supported by the discovery of frescoes on the cryptoporticus wall depicting a female figure resembling Artemis standing in a chariot (ibid.). Cryptoportici are known to have had cave-like functions in some cases, notably for incubation and healing as at the Asklepion at Pergamon. A similar case is found in the “den” of a group of priest-healers known as pholarchs at Elea (Ustinova 2009, pp. 191-209; Rickert, 2014, pp.477-479) [15]. This cult, of which the philosopher and poet Parmenides was a member, was dedicated to Apollo [16]. In an interesting parallel to the activities that may have occurred at Hylai, the priests of Apollo at Elea (called Hyele in Herodotus) went into incubation not to heal themselves, but to receive divine wisdom. If these two examples are any clue, we may imagine that the cryptoporticus at Magnesia functioned as a place of restricted access where priests or other functionaries withdrew from the crowds of the Agora into a dark, silent cave-like space where they could prepare themselves mentally to officiate in rituals.

Nearby in the agora, Zeus was worshiped in his chthonic guise ‘Sosipolis’ and honored with the sacrifice of a bull as befits an underworld deity (Kern 1900, no.98) [17]. The orientation of the temple of Zeus Sosipolis, which faces west, may be evidence of the architectural emphasis of his chthonic nature (Bingöl 2007, 112). Zeus is one of two prominent deities at Magnesia who were said to have been born, or raised, in a cave [18]. The other is Dionysos, whose characteristic representation on Roman coins of Magnesia as a child seated on a mystical cista mystica, a basket containing a serpent used in ritual, also connotes the underworld (Robert, 1997, Fig.10, p.67; Schulz, 1975, p.38) [19]. It is suggestive of a particular attunement to the functions of the cave that both of these gods were represented at Magnesia in a guise which emphasized these chthonic aspects of their nature. The origins of Zeus and Dionysos recall the cave as a nurturing place, which provided protection and seclusion. The presence of the cryptoporticus in the Agora may evoke the healing aspects of the cave. These benevolent properties are balanced by the cave’s function as a gate to the underworld, as evidenced by the cycle of death and rebirth symbolized in the sacrifice of the bull to Zeus Sosipolis.

Today, the underworld still has a presence at Magnesia, albeit a less allusive one. Bingöl (2007, p.163, p.178) recounts that underground passages persist in the imagination of the local population, although this is certainly due to the fact that half-buried structural vaults have the air of secret tunnels. One of these is the vaulted substructure of the Roman temple north of the Lethaios, which according to local folklore is said to be the start of an underground passage leading to the gymnasium. This explanation is probably more colorful than is warranted. The dimensions of the vaulted space correspond closely to the space supporting the podium of the temple of Augustus at Antioch in Pisidia. Mitchell and Waelkens (1998, pp.119-120, p.158) convincingly demonstrate that the vault under the temple at Pisidian Antioch, like many other temples dedicated to the imperial cult [20], was a byproduct of the desire for an elevated podium, that perhaps served as a treasury or other cellar, instead of a cult room intended to emulate a cave, as is the case with the Temple of Zeus at Aezani.

Modeling the Space of Ritual

The cults dedicated to Apollo at Hylai and Dionysos in the city interacted with the urban fabric of Magnesia on the Maeander through the dynamic of movement. This movement passed from the mountain slopes of Thorax, a sacred natural space which long functioned as a topographic anchor in the memory of Magnesians, through the city walls and, the processional route taking its shape from the planned grid of the streets, interacted with the architectural space of the city. The dendrophoroi, in bearing their fully-grown trees, intermingled the natural space of which their cave was the spatial apex, with the rationally-planned built environment of the Greek polis. In order to more fully understand how this movement and ritual functioned as “operative polytheism” (to paraphrase Detienne), I elaborate this performative dynamic within an architectural reading of the city. Specifically, an experimental method applied through the use of 3D models will shed light on how the architecture and planning of Magnesia responded to this ritualistic context.

The Path of the Dendrophoroi

The cave of Apollo at Hylai, if located in the vicinity of the site in the foothills of Thorax that directly overlooks Magnesia, would have been connected most directly with the city by way of a road that extended more or less directly down the ridge, traversing the city gate at the point directly above the stadium where the remains of a tower stand. In fact, a modern road (which ultimately diverts towards the village of Argavlı in the south rather than towards the ancient city) still follows a similar route today (Fig.35). The prominence of this slope as seen from the city suggests that a procession traveling down this road from the mountain may even have been visible from the city below, as the model seems to indicate [21].

View from the stadium of Hylai and the proposed route of the dendrophoroi

After passing the city gate at the tower, the dendrophoroi could have continued their descent down one of the hills that flank the stadium, which would have provided convenient seating for spectators. Perhaps they proceeded through the stadium itself (Fig.35, route b), though further excavation of the upper tiers of the sphendone would have to verify the feasibility of this. However they reached the foot of the stadium hill, from there it is less than 300 meters to the site of the presumed temple of Dionysos. If the procession took the road east of the stadium (Fig.35, route c), it may have traveled a road that passes by the conjectured site of the temple of Sarapis, and is on axis with both the city gymnasium and the temple to the north of the Lethaios. Possibly this area was less densely built up than the neighborhoods surrounding the city center, and thus more accommodating the transport of large trees and large crowds of spectators. As late as the mid-third century, if the inscription of the founding of the Dionysian cult is accurate, the site of the Dionysos temple was occupied by a grove of plane trees, suggesting that it was located in a quarter of the city that was undeveloped, or had ample room for gardens and orchards.

View from the stadium of dendrophoroi approaching the temple of Dionysos

The position on a major east-west route of the finds that Humann identified as Dionysus’ temple seems to indicate that the temple would have been sited on the street, oriented to the south, with a clear view of the tower and gate in the city wall beyond the stadium whence the dendrophoroi may have emerged during their descent into the city. As is the case of the alignment of the Artemis temple, here the question is not of human visibility as much as symbolic orientation in space. If the Dionysus temple did face south, it did so in distinction to Magnesia’s other major temples (e.g. Zeus and Artemis), which faced west or south-west. The north-south alignment of the Dionyos temple would underscore the temple’s urban embeddedness with the street grid, its character of being kata ten polin, “of the city”.

Possible paths of the dendrophoroi into the city

Subverting the Grid: The Polytheistic Network

All of this activity occurs in the south-east quarter of the city, staying clear of the civic center represented by the agora, prytaneion, market basilica, and odeon, as well as the sanctuary dedicated to Artemis Leukophryne. In fact, when spatially mapped onto the topography and architecture of the city, the Apollonian-Dionysian duality seems transformed into a triad. The position of the temple of Dionysos acts as a mediator or buffer between the two facing poles of Apollo and Artemis – the siblings distinguished here by their diametrical opposition: one high, one low; one dwelling within a cave surrounded forest, one residing in a magnificent temple surrounded by stoas; one outside the city’s walls, the other within its heart. From this perspective, the two seem to face each other, and between them lies the city of Magnesia, watched over by both gods. The 3D model shows that from the pronaos of the temple of Artemis, looking directly forward, one would have had a clear view of the site in the foothills of Thorax perhaps called Hylai, even with the altar of Artemis interposed. A line directly perpendicular to the front of the temple would bypass the site of ‘Hylai’ to the north, and in fact also misses the actual summit of Thorax. The orientation of the temple of Artemis towards the triangular peak where Hylai sits is therefore an illusion, yet visually the heights of Hylai dominate the skyline as seen from Magnesia, and it was probably perceived as part of the mountain, a visual stand-in for the actual, hidden peak. Though the cave is hidden, the mountain is visually present from every point in the city. The prominent visual framing of city as seen from the hilltop perch of Hylai completes the enfolding of the city within the purview of its protective deities.

The pediment of the temple of Artemis echoes the shape of Mt.Thorax

Reconstruction of temple facing the mountain.

In the center of the city’s grid, Dionysos stands between the two siblings. The axis of the temple of Artemis, although bypassing Thorax/Hylai, precisely intersects the location proposed by Humann for the Dionysos temple. Whether or not this was intentional, it does indicate that the spatial function of the temple should be viewed with consideration of its centrality within the city. The location of the thiasoi may give us further clues to Dionysos’ role in the city’s sacred polarities. As suggested above, the thiasos of Kosko, platanistinoi, might be sought at the location of the temple itself, where the plane tree of the original epiphany once stood. The second maenad, Baubo, is perhaps to be sought near the theatron, where the spring which sits above the fountain in the agora might indicate the location of Tabarnis. If the third thiasos of the maenad Thettale, the kataivatai, is in fact associated with the decent into a cave such as that of Apollo, then this might have been located in the village of Hylai. Together these form an elongated triangle that frames the stadium, and points to but does not penetrate the civic/religious core of the agora/Artemision. From this perspective, the dispersal of the cult of Dionysos in three thiasoi seems a strategic deployment. A triangular zone is created, which ‘contains’ the wild and irrational behavior of the dendrophoroi so that it cannot disturb the rest of the city. The stadium, at the center of this triangle, may have been a focus of the activity.

Network of Dionysian thiasoi

Locating the Cave

A central goal of this study has been to locate the cave of Apollo at Hylai both physically and conceptually within Magnesia’s urban topography. Yet the fact that the actual cave has never been found may be beside the point in some ways. Caves are by nature hidden things, not visible monuments but secret places buried within the earth. Very few of the caves mentioned in ancient sources have been located, unless they were situated beneath or adjacent to a monumental building such as a temple that attracted archaeological attention. Given that the cave for the Greeks was a natural cave, usually associated with springs of water and the bees which were considered to be sacred to nymphs, the indigenous traditions of rock-cut architecture the first Greek settlers encountered in Anatolia, notably in Lycia and Phrygia, probably had little influence on the Greek idea of “caveness”. Anatolian rock-carvers created meticulously articulated facades, but the interiors of these monuments, mostly intended as tombs, were dark and small. By contrast, while Greek architecture has sometimes been characterized by its externality (Zevi, 1957, pp.76-78), the cave for them was all about interiority, descending into darkness, going within [22]. It thus remained distinctly non-architectural, though never completely disassociated from architecture and civilization, but rather in dialogue with it.

Based on the combined testimony of the accounts of Pausanias and Damascius, and the representation of dendrophoroi on the city’s coins, we may safely assume that citizens of Magnesia would have been aware of the famous sanctuary of Apollo at Hylai, and that some individuals made pilgrimages there. Yet most residents probably did not know its exact location. The idea of the cave was in many ways more important than the physical presence of the cave itself. In relation to the city, it was a place both underground and high up, correlating to both sides of Apollo’s nature. Spectators watching the approach of the dendrophoroi may have associated the ritual as much with Dionysos as with Apollo, because of the culmination of the journey at Dionysos’ temple, and because the behavior of the dendrophoroi corresponds more to typical Dionysian ritual. The cave of Hylai, then, both in terms of its physical location and its associated practices, is hidden. It is unlikely that the place was on a regular pilgrimage route, because in contrast with most caves associated with Apollo, Hylai does not seem to have been an oracular site. The Magnesians were devoted to Pythian Apollo, traveling repeatedly all the way to Delphi to consult the oracle there. The existence of a local competitor would seem to be in conflict with this tradition. Nor does Hylai appear to have been a charonium, in the manner of Acharaca, Thymbria, or Hierapolis. Magnesia’s necropolis lies in the plain well below the heights of Thorax, suggesting that the “steep precipices” and forests with large trees where the gardeners of Apollo dwelt were free of funerary associations. A possible connection with these caves may be that the geological activity that produced mephitic vapors in the charonia of the Maeander valley had an intoxicating effect that caused the “strength equal to any task” that inspired the dendrophoroi. However, in this case the devotees of Apollo were not inspired to prophecy, or healing as happened in the rite at the Plutonium of Acharaca. In the reverse of the procession of the Nyseans towards Acharaca, movement started at Hylai and was directed towards the city of Magnesia, with the ultimate goal of the temple of Dionysos. The cave of Apollo was therefore implicit in the three spatially dispersed outposts of this cult, the Dionysian thiasoi. Though visually and physically inaccessible, the cave was present in the city, permeating its rationalized grid by means of ritual movement and spectacle.

It makes sense that Dionysos was the vehicle chosen for this infiltration by Apollo himself, if we recall the oracle that instituted the cult at Magnesia. Apollo had been at Hylai long before the founding of the city, yet he had no way of entering it, until Dionysos made his appearance and the maenads were installed. Pythian Apollo would naturally work in conjunction with Dionysos, with whom he shared the precinct of Delphi and the Corycian cave. The Magnesians, loyal to Apollo Pythios for his role in leading them to their new city, may have felt compelled to honor him by incorporating the ancient and strange cave of Hylai into their civic identity. This was not accomplished though architecture, since the cave was probably in too remote and precipitous a site to accommodate the building of a temple. Rather, they built a temple to Apollo’s proxy Dionysos in the center of the city, at the spot which the epiphany helpfully indicated.

It is a unique solution but not entirely without context. Sacred caves abounded in Greece, including many that were associated with buildings and cities, but there was no single dominant architectural strategy for dealing with them. This was surely due in part to the inherent formlessness of caves themselves, which are part of nature and thus resisted humanizing classicism. Yet as Nilsson (1961) shows, the polytheistic meaning with which the natural world was imbued was not incompatible with design strategies meant to evoke a rational and democratic society. Rather, the two formed two overlapping matrices. As the experiments of the previous chapter show, the type of grid represented by Hippodamian planning is readily represented with line drawings, arithmetical calculations, and diagrams. The second type, however, more readily responds to the expanded analytical mode of 3D modeling. Reductive drawings showing axes and paths do not adequately visualize the polytheistic grid; this type of network was tied to the natural terrain and perceived from eye level as a human perceives the landscape, a far cry from the totalizing bird’s-eye view implied by the urban grid that is the purview of gods or those with god-like powers.


[1] English translation from Briant (2002). Original published in Cousin and Deschamps (1889).
[2] This passage, as well as all the following quotations from Texier, Rayet, Kern, and Robert, are my translation.
[3] By way of comparison, the Nyseans carried a live bull 2km to the Plutonium at Acharaca (Strabo 14.1.44), but this route was relatively level throughout. Apart from the greater distance, the south bank of the Maeander represents a conceptual difference – it was an unsettled area far from the well-travelled route that connected Ephesus, Magnesia and Tralles, whereas Acharaca and Nysa were linked by topography and roads.
[4] In this text, Apollonos Aulai is named, instead of Hylai. Henceforth in the text it is assumed that the two are synonymous and ‘Hylai’ will be used for the sake of simplicity.
[5] Transcribed in Photius, extract 116.
[6] Rather than Leukophrys, as conjectured by Phillipson (1936).
[7] See also Reinach (1890) and Kern (1895). These discussions are updated by Henrichs (1978) and Graf (2004).
[8] Translated by Henrichs (1978).
[9] A distinction should be made between the names of the thiasoi, which describe groups of people and the names of the maenads’ burial sites, which refer to places. However, if the maenads were buried in separate, significant locations it seems likely they may have practiced in different neighborhoods as well. The names of their thiasoi may well be clues, as the spatial designation pro poleos, for example, appears to indicate.
[10] Thonemann (2011, p.257) however, considers Tabarnis to be an “indigenous, or at least very ancient Greek” name.
[11] See also Maass (1891,186)
[12] According to Callimachus (Hymn to Apollo, verses 55-64) Apollo “delights in the founding of cities” and was therefore called Archegetes (leader and protector of colonies).
[13] On origins and sources for the story of Apollo and the Hyperboreans, see Fontenrose (1959,pp.382 – 383; note 25 p.382)
[14] Ustinova (2009, 121-155) gives an extended account of the controversy surrounding the Delphic adyton.
[15] The term pholeos was also used by Strabo in his description of the cave of Archaraca near Nysa in Caria, which also had healing properties (Strabo, 14.1.44; Ustinova, 2009, p.198). The connection between the two practices is supported by the fact that Elea was a colony of the Phocaeans, who fled Caria in the Persian era (Herodotus 1.167).
[16] Parmenides’ famous poem, “On Being”, essentially enacts the reverse of Plato’s allegory, in that it describes a katabasis, or descent to wisdom and truth (Rickert, 2014).
[17] Zeus Sosipolis was considered a chthonic god (Long, 1987, pp.248-9). The procession of youths bearing the sacrificial bull is reminiscent of the procession from to the Plutonium at Acharaca (Strabo 14.1.44)
[18] Zeus was hidden in a cave under Mt.Ida on Crete as a baby, to protect him from his father Cronos, who believed he was destined to be killed by his son. Caves sacred to Zeus were common Greece (Weinberg 1986, pp.115-118). It is toward the cave of Zeus on Crete that the group of philosophers is walking as they describe the ideal city of Magnesia in Plato’s Laws.
[19] The child Dionysos was raised by Nymphs in a cave in the idyllic peninsula of Nysa (Weinberg 1986, pp.120-124).
[20] Cf. the Maison Carrée at Nimes or the temple of Rome and Augustus at Caesarea Maritima (Amy, R. and Gros, P. (1979). La Maison Carrée de Nîmes. Éditions du Centre National de la Recherche Scientifique).
[21] On the ancient road, see Keil, (1908, pp.166-7) and Thonemann (2011, pp. 103 – 104). The road is shown in the map by Phillipson (1936)
[22] The Romans, who like the Anatolians came from a landscape of soft tufa and also had a tradition of rock-carving dating back to the Etruscans, combined aspects of interior space and exterior monumentality in their cave-spaces.