Hymnus invicto:

The structure of Mithraic cult images with multiple panels

von ALEXANDER KIRICHENKO, Cambridge, MA

The meaning of the Mithraic tauroctone reliefs and frescoes has puzzled
scholars for over a century now. The amount of literature dedicated to Mi-
thraic art in general or to various aspects of particular monuments is truly
overwhelming. Each concrete detail of Mithraic cult images tends to generate a
great number of diverging opinions often bordering on wild speculation1. This
interpretive crux arises, in part, from the fact that we do not have much to go
on besides the images. In other words, there is no extant literary source that
would help us to produce a plausible interpretation of the significance attri-
buted by Mithraists to the complex visual representations of their god. Survi-
ving references to the Mithraic cult in Porphyry2 and early Christian writers3
are, unfortunately, too inconclusive to be of much help.

It is beyond doubt that Mithraic art is laden with complex symbolism. The
image of Mithras slaying the bull, which was a focal point of any Mithraic
temple (mithraeum) from Syria to Britain, invariably contains a number of ele-
ments that urge the viewer to interpret the whole scene as a mythical and
symbolic representation of the creation of the existing world order. Mithras,
who is himself identified with the Invincible Sun (Sol Invictus) by innumerable
inscriptions4, often wears a cloak studded with stars. Figures of Sol and Luna
often appear at the two upper corners of the cult image. Depictions of the four
winds, the twelve planetary gods and/or the twelve Zodiac signs are also
quite frequent. On all surviving tauroctone images, the bull’s tail turns into
ears of corn, and various animals – a scorpion, a snake, a dog, and a lion –
cling to his body partaking of his blood and semen. In most of the images, the
figure of Mithras is flanked by two torchbearers – one raising, the other lowe-
ring the torch – whose names are known from inscriptions to be Cautes and
1
Here are some of the most significant book-length studies of the Mithraic mysteries: F.
Cumont, Les mystères de Mithra, Brussels 1902; M.J. Vermaseren, Mithras, the Secret God,
London 1963; A. Schütze, Mithras. Mysterien und Urchristentum, Stuttgart 1972; R. Merkel-
bach, Mithras, Königstein 1984; P. Ulansey, The Origins of the Mithraic Mysteries. Cosmology
and Salvation in the Ancient World, Oxford 1989; R. Turcan, Mithra et le Mithraïcisme, Paris
1993; M. Clauss, The Roman Cult of Mithras, New York 2000.
2
Porphyry, The cave of the nymphs in the Odyssey; a revised text with translation by Seminar Clas-
sics 609, State University of New York at Buffalo (= Arethusa Monographs 1), Buffalo 1969.
3
E.g., A.S. Geden, Mithraic sources in English, Hastings 1990.
4
M.J. Vermaseren, Corpus Inscriptionum et Monumentorum Religionis Mithriacae, The Hague
1956-60 (henceforth CIMRM) passim.

Göttinger Forum für Altertumswissenschaft 8 (2005) 1-15

http://www.gfa.d-r.de/dr,gfa,008,2005,a,01.pdf
2 Alexander Kirichenko

Cautopates. Their respective positions to the left and to the right of the bull-
slaying god often (but not always) correspond to the placement of the anthro-
pomorphic figures of Sol and Luna at the upper left and right corners. Does
Cautes then symbolize light and Cautopates darkness, as many have sugge-
sted?5 Or are Cautes and Cautopates two manifestations of the solar god Mi-
thras himself – the rising and the setting sun, as seems to be indicated by some
inscriptions?6 Or are they simply mythical companions of Mithras assisting
him in his deed and imbued with some other significance? Or can all these
possibilities be reconciled? Similar questions can be asked about any of the
elements of the Mithraic cult image that I have listed above. Depending on
how one answers these questions, one can reconstruct the myth of Mithras,
along with its theological implications, in a variety of ways. However, in the
absence of any relevant textual material, any such reconstruction would have
to remain purely arbitrary.

Some of the most elaborate Mithraic cult-images have the central panel sur-
rounded by a series of smaller ones. There have been a number of attempts to
extract a coherent story from these complex paneled Mithraic reliefs – each of
them implemented with a certain degree of plausibility, in spite of the lack of
reliable external evidence. The idea underlying most of these reconstructions,
namely that one can fruitfully employ Iranian material as a key to many pro-
blems posed by Mithraic imagery, has been generally abandoned and replaced
by a much more sober view that the Mithraic mysteries are a product of the
late antique syncretistic mind, with Rome being their most probable birth
place7. A more recent tendency to read Mithraic images against the back-
ground of Greek astrology8 or Neo-Platonism9 has been quite successful in
elucidating some possible connections between the mysteries of Mithras and
their cultural environment, although most conclusions in this field remain pu-
rely speculative.

5
E.g., R.L. Beck, Cautes and Cautopates: some astronomical considerations, JMS 2, 1, 1977, pp. 1-17.
6
CIMRM 836 = CIL VII 1344c Deo M(ithrae) / C(auto)p(ati) S(oli) I(nvicto); Vermaseren, Mi-
thras (see note 1), pp. 72-73.
7
A.D. Nock, The Genius of Mithraism, JRS 27, 1937, pp. 108-114, repr. in his Essays on Reli-
gion and the Ancient World, ed. Z. Stewart, Oxford 1972, pp. 452-458. See also M.P. Nilsson,
Geschichte der griechischen Religion, Bd. 2, Munich 1966, and M. Clauss, The Roman Cult of
Mithras, New York 2000. The idea of the Persian origin of the Roman mysteries of Mithras
has been resuscitated in a modified form by R.L. Beck, The Mysteries of Mithras: a new ac-
count of their genesis, JRS 88, 1988, pp. 115-128.
8
E.g. R.L. Beck, Planetary Gods and Planetary Orders in the Mysteries of Mithras, EPRO 109,
Leiden 1988.
9
R. Turcan, Mithras Platonicus: recherches sur l’hellénisation philosophique de Mithra, EPRO 47,
Leiden 1975. See also A. Blomart, Mithra et Porphyre: Quand sculpture et philosophie se re-
joignent, Revue de l’Histoire des Religions CCXI-4, 1994, pp. 419-441.
 Hymnus invicto 3

Mithraic art is indeed mind-boggling and strange, and scholars will probably
never stop trying to decipher the story that it tells, no matter how unattainable
this goal might be. What I think can potentially produce more tangible results,
which would at the same time be highly relevant to our overall understanding
of the late antique religious mentality, is not the question, ‘What is the central
myth of Mithraic mysteries and how is it related to doctrines and rituals of the
cult?’ but rather, ‘Do the Mithraic images with multiple scenes represent a
story at all and, if so, how do they represent it? Do we deal with a pictorial
narrative here or with a completely different kind of visual discourse?’ An at-
tempt to answer these questions can help us better understand the function of
the tauroctone image in Mithraic rituals.

The intuitive reaction of any viewer looking at the side panels of Mithraic cult
images would be to interpret them as representing some sort of mythical nar-
rative about the deeds of Mithras. The analogies are close at hand: vases repre-
senting deeds of Theseus and Heracles, and, even more closely, the Tabulae
Iliacae10. Cumont11 was the first to interpret the side panels as a narrative by di-
stinguishing several major episodes that appear on many of Mithraic reliefs
and frescoes. The way he reconstructed the succession of the individual episo-
des sounds quite plausible. A number of them depict the stage preceding the
birth of Mithras: one scene showing a reclining figure tentatively identified by
Cumont as Saturn12, another scene supposed to represent Jupiter fighting
against the giants13. The next stage is the birth of Mithras from the rock14. What
follows are various exploits accomplished by Mithras. One of them is the so-
called ‘water-miracle’, where Mithras produces water out of the rock of a cave
by shooting an arrow against it15. The next episode shows Mithras’ encounter
with the bull. This episode takes up a few panels: we see the bull peacefully
grazing16, Mithras on the bull’s back17, the bull on Mithras’ back18, the Torch-
bearers carrying the dead bull19. The culmination of this series – the killing of
the bull – is represented on the central panel. The next stage of the story is the
encounter between Mithras and the Sun (Sol). It often consists of a number of
panels as well: Sol kneeling before Mithras who holds the bull’s haunch20, Sol
10
See the parallels in F. Saxl, Mithras. Typengeschichtliche Untersuchungen, Berlin 1931.
11
F. Cumont (see note 1), pp. 104-149.
12
CIMRM 42, 194, 390, 650, 723, 966, 1083, 1283, 1292, 1359, 1400, 1430.
13
CIMRM 42, 194, 390, 650, 723, 966, 1083, 1292, 1359, 1400, 1430.
14
CIMRM 42, 194, 390, 650, 966, 1083, 1161, 1247, 1283, 1292, 1301, 1400, 1430.
15
CIMRM 42, 194, 390, 966, 1083, 1225, 1283, 1292, 1294, 1301, 1359, 1400, 1430.
16
CIMRM 1283, 1292.
17
CIMRM 42, 194, 966, 1137, 1247, 1283, 1292, 1400.
18
CIMRM 42, 1137, 1283.
19
CIMRM 42.
20
CIMRM 42, 194, 390, 650, 729, 1083, 1137, 1400, 1579.
4 Alexander Kirichenko

and Mithras standing by an altar21, Mithras and Sol shaking hands (dextrarum
iunctio)22, Mithras and Sol feasting together23, Mithras entering Sol’s chariot24.

It is quite likely that the story known to practicing Mithraists ran along similar
lines. However one has to be aware of the fact that most of the identifications
proposed by Cumont and later accepted by other scholars are tentative and
can in many cases be subjected to doubt. Thus, as I pointed out at the begin-
ning, it still remains unsolved what exactly is depicted on each of the side pa-
nels of Mithraic cult images.

Side panels pose other questions as well. Cumont paid almost no attention to
the arrangement of the individual side panels on the surviving monuments
nor did he give any explanation of the numerous scenes that do not fit into the
narrative frame. If we consider the position of the same scene on various mo-
numents, we will see that it varies considerably. Some scenes tend to occur
next to each other, thus forming short narrative sequences (e.g. Mithras and
the bull, Mithras and Sol), but the arrangement of the whole is never the same.
Furthermore, on some cult images thematically related panels are separated
by incongruous episodes that interrupt the apparent narrative flow. Additio-
nally, there are no monuments that have exactly the same selection of scenes.
The number of overlapping scenes is relatively great, but all in all there is only
one scene that occurs on all surviving monuments25.

There have been a number of attempts to explain this situation. Most of them
use Cumont’s assumption as their starting point: what is represented on the
side panels is a series of illustrations of an epic style narrative relating the
deeds of Mithras, whose primary purpose was to instruct the neophytes un-
familiar with the story. Saxl26 assumed that each individual extant monument
represents a certain degree of deviation from the original prototype that con-
tained all episodes of the narrative that were supposed to be read in a certain
direction. He thought that this massive distortion occurred as the cult migra-
ted from its birthplace in Asia Minor to the West. This theory is clearly unte-
nable both because very few people now believe in the origin of the cult in

21
CIMRM 390, 650, 1510, 1579.
22
CIMRM 194, 650, 1292, 1400, 1430.
23
CIMRM 42, 390, 966, 1137, 1247, 1292, 1400.
24
CIMRM 390, 1083, 1247, 1283, 1292, 1294, 1400, 1430, 1579.
25
R. Gordon, Panelled Complications, JMS 3, nos. 1-2. Henley on Thames 1980, repr. in his
Image and Value in the Graeco-Roman World: Studies in Mithraism and Religious Art, Alde-
rshot 1996, diagrams on pp. 212-218.
26
F. Saxl (see note 10).
 Hymnus invicto 5

Asia Minor and because the discrepancy between different monuments is too
great to allow for such a reductive explanation.

E. Will27 tried to distinguish several types of Mithraic reliefs, each one charac-
terized by a particular reading order. He postulated that each of them was ty-
pical for a certain geographical area. However, he had to admit that this
hypothesis did not account for a great number of exceptions that did not fol-
low the imposed narrative sequence.

R. Gordon28 analyzed all of the completely or partially extant Mithraic reliefs

with multiple panels and concluded that they do not represent any narrative
sequence at all. He refuted Saxl’s argument that the original narrative order
was distorted during transmission, and noted that it would be disrespectful
for us to think that ancient initiates did not know the sense of their own my-
steries. He pointed out that, despite the fact that there is a certain amount of
regularity in the way some of the individual scenes are arranged, it is still
virtually impossible to establish a particular reading order for most of the mo-
numents. Furthermore, he noted that some side panels do not fit into any nar-
rative structure at all and that some find themselves in completely different
surroundings on different monuments29. One would have to know the mea-
ning of each particular episode depicted on the side panels beforehand in or-
der to understand the meaning of the whole. According to Gordon, the side
panels must therefore have served a different purpose. He assumes that they
might represent particular parts of the ritual performed in the mithraeum or
illustrate certain aspects of the Mithraic theology. He tries to corroborate this
point by referring to the fact that one of the panels depicts a banquet which is
also represented on the back of some of the tauroctone reliefs where it clearly
refers to a ritual rather than a mythical scene. He also allowed for a possibility
of a deeper symbolic meaning of the seemingly disorderly arrangement, since
most of the scenes occur either on the left or on the right side, which would
clearly correspond to the Mithraists’ obsession with the orientation of their
temples. The last supposition has later been refuted by S. Zwirn30, who has
pointed out that there are a few examples of the same scene represented on
either side. Additionally, even the positions of the two torchbearers Cautes
and Cautopates, who are thought to symbolize light and darkness respecti-
vely, are reversed on some of the cult images31.

27
E. Will, Le relief cultuel gréco-romain, Paris 1955.
28
R. Gordon (see note 25).
29
Ibid. See diagrams on pp. 212-218.
30
S. Zwirn, The Intention of Biographical Narration on Mithraic Cult Images, Word and Image 5,
1989, pp. 2-18.
31
E.g., reliefs from Rome (CIMRM 350), Bononia (CIMRM 693), and Nida (CIMRM 1083).
6 Alexander Kirichenko

R. Merkelbach32 thinks that the side panels have a ritual significance too. He
points out that some of the scenes represented on the side panels also occur as
free-standing statues, from which he concludes that all these scenes may in
some way be symbols of different grades of initiation. According to Merkel-
bach, all human figures depicted on the side panels are not gods, but rather
initiates performing various rituals. Merkelbach uses this assumption as a ba-
sis for a detailed reconstruction of what he thinks happened during the initia-
tion rites and of the symbolic significance individual actions may have had.
This approach obviously makes the very notion of narrative completely irrele-
vant for the discussion of the side panels, but it does not seem to offer any im-
provement instead, since the changing order of individual episodes still re-
mains unexplained.

S. Zwirn33 admits all the difficulties involved in analyzing Mithraic reliefs.

Nevertheless he returns to the notion of narrative as the primary purpose of
the side panels. He analyzes the complex Mithraic cult images in terms of the
viewers’ perception, and postulates that late antique viewers, familiar with
such monuments as funerary sarcophagi and triumphal arches, would inevita-
bly recognize a biographical intention in Mithraic art as well. This is probably
true, but it does not explain why Mithraic artists chose to represent these mul-
tiple images in this particular way and what motivated their choice of indivi-
dual episodes.

Each of the studies I have mentioned above has undoubtedly made an impor-
tant contribution to our understanding of how the side panels are arranged on
the surviving monuments. None of them, however, has offered a satisfactory
explanation of why they are arranged in the way they are. R. Gordon has, in
my opinion, quite conclusively demonstrated the intrinsic failure of any at-
tempt to interpret the side panels as a narrative sequence. His own and R.
Merkelbach’s attempts to interpret the side panels as scenes of ritual or illu-
strations of religious concepts could clearly be one way of dispensing with the
notion of narrative in the Mithraic reliefs. But their hypotheses are not sup-
ported by any conclusive evidence. Additionally, getting rid of the idea of nar-
rative should not necessarily imply that there has to be no rationale behind the
arrangement of the individual scenes. Unfortunately, Merkelbach completely
ignores this issue, while Gordon only offers a tentative explanation that does
not seem to hold true under further scrutiny.

32
R. Merkelbach (see note 1), pp. 75-132.
33
S. Zwirn (see note 30), p. 8.
 Hymnus invicto 7

Another weakness of this approach is that it completely rejects the idea that
the side panels reflect a certain mythical story. However, the fact that the bull-
slaying scene at the center of the Mithraic cult image is thematically and styli-
stically linked with other panels depicting interactions between Mithras and
the bull suggests that the tauroctone scene is the culmination of the events on
the side panels. There is also no alternative to the interpretation of a male fi-
gure emerging from the rock as an episode of Mithras’ mythical biography,
since he is known from some inscriptions to have been born from the petra ge-
netrix.34

Thus, it appears to me that the side panels, along with the central tauroctone
icon, depict different episodes of the Mithraic myth, but they depict them in a
non-narrative order. Nevertheless I do not think that the arrangement of the
side panels is simply done at random and is not governed by any reason at all.
In order to uncover the logic behind this seemingly irrational arrangement, we
have to ask ourselves whether there is any other form of religious art, be it vi-
sual or literary, that would be characterized by the same disregard for chro-
nological cohesiveness.

Many interpretations of the side panels have been based on the assumption
that they represent a canonic written narrative, which also played a certain
role in the Mithraic ritual. Cumont was absolutely sure that there existed a
vast corpus of Mithraic literature and deplored its irretrievable loss. It seems
likely, however, that this idea is not based on anything but a deep-rooted un-
critical assumption that most of the nineteenth century historians of ancient
religion never subjected to doubt. Much of the ancient evidence was viewed,
as it were, through the prism of a modern interpretatio Christiana that was una-
ble to detach itself from the familiar contemporary forms of religion. One of
the most blatant examples of this tendency is the treatment of ‘Orphism’ in 19th
and early 20th-century handbooks of ancient religion. It was invariably repre-
sented as the Orphic ‘church’ relying on a set of canonic sacred writings often
called the ‘Orphic Bible’35. Cumont seems to interpret Mithraic mysteries along
similar lines. Later, however, it has been repeatedly shown that this Christian
model contradicts other knowledge about ancient mystery cults36. Burkert
emphasizes the fact that mystery doctrines were kept secret from the unini-
tiated and were most likely transmitted orally. Furthermore, from the texts

34
See CIMRM 1127.
35
For the criticism of this approach, see U. von Wilamowitz-Möllendorf, Der Glaube der
Hellenen, Bd. 2, Berlin 1932, pp. 192-204; I.M. Linfort, The Arts of Orpheus, Berkeley-Los
Angeles 1941.
36
W. Burkert, Ancient Mystery Cults, Cambridge/Mass.-London 1981, pp. 30-65.
8 Alexander Kirichenko

whose relation to a particular cult can be demonstrated we cannot draw the

conclusion that sacred narratives played a role in any mystery ritual. Thus,
what an ancient mystes would probably have known is the story that the cult is
based on rather than a sacred written narrative with a canonical status.

It does not mean that the mythical background and the doctrines of the myste-
ries were not reflected in any kind of text at all. Mystery initiation rituals
(often referred to in ancient sources as dromena) were accompanied by utteran-
ces (known as legomena). These were kept absolutely secret and did not sur-
vive therefore, except for a few references made by initiates who converted to
Christianity and did not feel obliged to keep the secret37. Texts that did survive
are some prayers and hymns originating from a few different contexts, which
either predate Mithraic monuments or are roughly contemporaneous with
them. I think instead of looking for answers to problems posited by Mithraic
reliefs in hypothetical sacred narratives, whose existence we cannot prove, we
could benefit much more from analyzing these surviving samples of religious
literature.

Ancient religion was very loosely organized and the boundaries between vari-
ous cults were very fluid. Apuleius, for instance, says in his Apology that he
was an adept of numerous mystery cults38, and a number of 4th century AD
inscriptions identify some devotees of Di Magni as priests of numerous other
gods including Mithras39. All these cults belonged to the same cultural milieu,
and they exerted a considerable amount of influence on each other or drew on
the same basic ideas. It seems to be quite legitimate therefore to use the mate-
rial of one such cult to elucidate certain basic features of another.

37
Ibid.
38
Apuleius, Apologia (= De Magia) 55: Sacrorum pleraque initia participavi. eorum quaedam
signa et monumenta tradita mihi a sacerdotibus sedulo conservo. nihil insolitum, nihil incognitum
dico. vel unius Liberi patris mystae qui adestis scitis, quid domi conditum celetis et absque omni-
bus profanis tacite veneremini. at ego, ut dixi, m u l t i i u g a s a c r a e t p l u r i m o s r i -
t u s e t v a r i a s c e r e m o n i a s studio veri et officio erga deos didici.
39
CIMRM 514 (376 AD): Diis magnis / Ulpius Egnatius Faventius / v(ir) c(larissimus) augur
pub(licus) p(opuli) r(omani) q(uiritium) pater et hieroceryx d(ei) s(olis) i(nvicti) M(ithrae) /
arch(i)bucolus dei Liberi / hierofanta Hecatae sa / cerdos Isidis percepto taurobolio criobolioque /
idibus augustis d(ominis) n (ostris) / Valente Aug(usto) V et Valentinia / no Aug(usto)
co(n)s(ulibus) feliciter / Vota Faventius bis deni suscipit orbis Ut mactet repetens aurata fronte
bicornes. Cf. CIMRM 513, 515, 516, 520, 522, 523, 524.
 Hymnus invicto 9

The most significant collection of ancient hymns related to a religious cult (as
opposed to purely literary hymns) are the so-called Orphic Hymns40. This is a
collection of 87 relatively short hymns addressed to a great number of deities.
There has been a long debate as to the possible date of the collection with sug-
gestions ranging from the 1st century BC up to the early Byzantine period. The
most commonly accepted period now is the 1st century AD. It is likely, how-
ever, that these hymns incorporate much older material. Since some of the rare
epithets attested in the hymns are also mentioned in inscriptions from Asia
Minor, it has been assumed that the collection as a whole originates from
there. The cult the Hymns originate from has not been clearly identified, but
since they contain numerous mentions of mystai we can assume that it was a
mystery cult. The hymns are characterized by henotheistic/pantheistic ten-
dencies and identify a lot of different gods with each other. Each of the hymns
consists of an invocation, a central part, and a final prayer. The central part is
normally just a series of epithets sometimes expanded into participial con-
structions or relative clauses. Some of the epithets epitomize, as it were, ex-
cerpts from a certain mythical narrative. Thus, behind each individual hymn
there are a number of mythical stories, but their purpose is not to tell these
stories, but to praise the gods. Therefore the order in which the epithets are ar-
ranged is anything but chronological.

Let us take, for instance, Hymn 29 to Persephone:

FersefÒnh, yÊgater megãlou DiÒw, §ly°, mãkaira,

mounog°neia yeã, kaxarism°na dÉ flerå d°jai,
PloÊtvnow polÊtime dãmar, kednÆ, biod«ti,
 kat°xeiw ÉA¤dao pÊlaw ÍpÚ keÊyea ga¤hw,
Prajid¤kh, §ratoplÒkame, DhoËw yãlow ègnÒn, 5
EÈmen¤dvn gen°teira, Ípoxyon¤vn bas¤leia,
n ZeÁw érrÆtoisi gona›w teknsato koÊrhn,
m
ter §ribrem°tou polumÒrfou EÈboul
ow,
ÑVr«n sumpa¤ktreira, faesfÒre, églaÒmorfe,
semnÆ, pantokrãteira, kÒrh karpo›si brÊousa, 10
eÈfeggÆw, kerÒessa, mÒnh ynhto›si poyeinÆ,
efiarinÆ, leimvniãsin xa¤rousa pnoªsin,
flerÚn §kfa¤nousa d°maw blasto›w xlookãrpoiw,
èrpagima›a l°xh metopvrinå numfeuye›sa,
zvØ ka‹ yãnatow moÊnh ynhto›w polumÒxyoiw, 15

40
The most recent book on the Orphic Hymns is A.-F. Morand, Études sur les Hymnes or-
phiques, Leiden 2001. It has detailed discussions of most of the debated issues and a vast
bibliography.
10 Alexander Kirichenko

FersefÒnh: f°rbeiw går ae‹ ka‹ pãnta foneÊeiw.

klËyi, mãkaira yeã, karpoÁw dÉ énãpempÉ épÚ ga¤hw
efirÆn˙ yãllousa ka‹ ±pioxe¤rƒ Íge¤&
ka‹ b¤ƒ eÈÒlbƒ liparÚn g
raw katãgonti
prÚw sÚn x«ron, ênassa, ka‹ eÈdÊnaton PlÒutvna. 20

We learn enough about Artemis from this hymn to be able to name her essen-
tial qualities and attributes and to reconstruct some elements of her mythical
biography. These elements, however, are not arranged in a chronological
order. The hymn tells us a lot about Artemis’ genealogical affiliations, but refe-
rences to her parents and her offspring are scattered throughout the text
among other epithets that depict her unchanging qualities and cannot be ar-
ranged into any kind of temporal progression. Chronological order seems to
be largely irrelevant here. Moreover, individual episodes can be alluded to
more than once in different ways, for instance Artemis’ birth from Zeus in ll. 1
and 7 and her marriage to Pluto in ll. 3 and 14.

This form of address to gods is also attested in magic incantations known as

Magic hymns. These hymns are preserved on magical papyri from Egypt and
date from the Imperial period41. In contradistinction to the Orphic hymns,
whose primary goal is to secure the overall well-being of the initiates both in
their earthly existence and in the afterlife, the Magic hymns are designed for a
much more specific purpose of attracting or harming particular individuals
through the mediation of the gods that they invoke. Despite these obvious dif-
ferences, both groups of hymns present a similar structural arrangement of the
invocation. In the Magic hymns we also have long series of epithets combining
attributes of gods with references to myths.

Such hymns must have been much more widespread than it might appear
from the evidence originating from just two different cult contexts. There are
two anonymous epigrams in the Anthologia Palatina that replicate a similar
hymnal structure in a rather playful way, i.e. they arrange the epithets alpha-
betically.

9,524

M°lpvmen basil
a fileÊion, efirafithn,

èbrokÒmhn, égro›kon, éo¤dimon, églaÒmorfon,
BoivtÒn, brÒmion, bakxeÊtora, botruoxa¤thn,

41
Papyri Graecae Magicae: Die griechischen Zauberpapyri, hrsg. und übers. von K. Preisendanz,
2. Aufl. von A. Henrichs, Stuttgart 1973-74 (= PGM), vol. II, pp. 237-266.
 Hymnus invicto 11

ghyÒsunon, gonÒenta, gigantol°thn, gelÒvnta,

Diogen
,
…
Àrion, …mhstÆn, »re¤trofon, »res¤doupon.
M°lpvmen basil
a fileÊion, efirafithn.

9,525

ÑUmn°vmen Paiçna, m°gan yeÚn ÉApÒllvna,

êmbroton, églaÒmorfon, ékersekÒmhn, èbroxa¤thn,
briyÊnoon, basil
a, belessixar
, biodthn,
ghyÒsunon, gelÒvnta, gigantol°thn, glukÊyumon,
Diogen
,
…
»kÊpon, »kuep
, »kÊsporon, …residthn,
ÑUmn°vmen Paiçna, m°gan yeÚn ÉApÒllvna.

These epigrams reduce the disorderly arrangement of the epithets to absurdity

by superimposing an artificial order on them which is absolutely irrelevant to
the meaning. It really makes the impression of a parody, which would be
meaningful only if the original form were widespread enough to be well
known to a chance reader42.

Another set of texts that show the same kind of disregard for chronological
order comes from yet another mystery cult. These are the so-called Isis areta-
logies43. In their full form they are all transmitted on inscriptions, although
Diodorus Siculus quotes an abridged version of a similar text (I 27). The ear-
liest example of this text is the inscription from Maroneia dating from the 2nd
century BC44. It is a prose encomium of Isis praising her for various benefac-
tions that she has bestowed upon humanity. Here Isis is addressed in the se-
cond person. A much more common form of this text preserves the general
tone and the structure of the earliest example, but turns it into a first person
account told by Isis herself. These texts are often referred to in modern scho-
larship as ‘self-revelations’ of Isis. There are a number of surviving represen-
tatives of this group – the most complete one in prose as well as a number of

42
There are other literary adaptations of this form of hymnal poetry, for instance, Nonnos,
Dion., 40,369-410, 44,191-216; the end of book 1 of Statius’ Thebais (and probably elsewhere).
All these passages also identify numerous deities with each other and accidentally the name
of Mithras is mentioned in two of them (Nonnos, Dion. 40,399 and Statius, Theb. 1,720).
43
See discussion in F. Solmsen, Isis among the Greeks and Romans, Cambridge/Mass. 1979,
pp. 27-52.
44
Y. Grandjean, Une nouvelle arétalogie d’Isis à Maronée, EPRO 49, Leiden 1975.
12 Alexander Kirichenko

poetic adaptations of it. They all replicate more or less the same confused or-
der of the earliest extant monument while expanding or reducing the number
of Isis’ achievements. I would like to quote an excerpt from the longest and
most complete version of the text (Isis aretalogy from Cumae)45:

…
3 âIsiw §g efimi  tÊranniw pãshw xraw ka‹ §paideÊyhn ÍpÚ ÑErmoË
ka‹ grãmmata eron metå ÑErmoË, tã te flerå ka‹ dhmÒsia, ·na mØ
to›w aÈto›w pãnta grãfhtai.
4 ÉEg nÒmouw ényrpoiw §y°mhn, ka‹ §nomoy°thsa ì oÈye‹w dÊnatai
metaye›nai.
5 ÉEg efimi KrÒnou yugãthr presbutãth.
6 ÉEg efimi gunØ ka‹ édelfØ ÉOs¤ridow basil°vw.
7 ÉEg efimi  karpÚn ényrpoiw eÍroËsa.
8 ÉEg efimi mÆthr ÜVrou basil°vw.
9 ÉEg efimi  §n t“ toË KunÚw êstrƒ §pit°llousa.
10 ÉEg efimi  parå gunaij‹ yeÚw kaloum°nh.
11 ÉEmo‹ BoÊbastow pÒliw ”kodomÆyh.
12 ÉEg §xrisa g
n épÉ oÈranoË.
13 ÉEg êstrvn ıdoÁw ¶deija.
…
57 Xa›re A‡gupte yr°casã me.

Here we see a list of cosmogonic and civilizing activities accomplished by Isis

which are presented in an absolutely jumbled order. It would be natural to as-
sume that the separation of the earth from the sky (12) must have preceded the
invention of writing (3) or the founding of the city of Boubastos (11), etc.,
which is, however, not the order in which individual statements occur in Isis’
self-revelations. They are all presented as a sequence of timeless events with
an equal significance ascribed to each of them. It would be a clear folly to as-
sume that the surviving Isis aretalogies distort the order of a more logical se-
quence of some hypothetical prototype. I think one can be fairly certain that
there never was a written text representing Isis’ activities in the form of a nar-
rative chronologically arranged. The lack of order here appears to be delibe-
rate and quite appropriate to the main objective of the text, i.e. the praise of
Isis. Nevertheless one can see that behind the lack of order there is an appro-
ximation of a notional narrative one could construct out of individual state-
ments by re-arranging them.

45
M. Totti, Ausgewählte Texte der Isis- und Sarapis-Religion, Hildesheim-New York 1985, pp. 1-4.
 Hymnus invicto 13

Both the relative chronology of the Isis texts and their similarity to the Orphic
Hymns suggest that the second person form preceded the self-revelation for-
mat. Thus we are dealing here with another example of an essentially hymnal
text originating from yet another cult context which displays a similar kind of
non-narrative structure as the Orphic Hymns.

The conclusion I would like to draw from this survey of some of the surviving
ancient hymns and related texts is that the confused order of the side panels
on the Mithraic reliefs might have been inspired by similar texts. Just as the
Orphic hymns addressed to the same deity can freely draw on a vast number
of traditional epithets incorporating as many or as few of them as seemed ap-
propriate, Mithraic reliefs are also very inconsistent in their choice of indivi-
dual scenes. Just as in some of the Orphic Hymns simple epithets can be re-
placed by much longer participial phrases or relative clauses containing some-
what more detailed information, individual episodes in the Mithraic reliefs
can be expanded to occupy a number of panels (Mithras and the bull, Mithras
and Sol). Just as in the Orphic Hymns we have numerous epithets describing
attributes and qualities of gods rather than episodes from their mythical bio-
graphies, some of the Mithraic reliefs contain panels that could not fit into any
kind of narrative structure at all46. Just as in the Orphic Hymns the same epi-
thet can be repeated within the same hymn, there are repeated scenes on some
of the Mithraic reliefs47.

We do not have any extensive hymns from the Mithraic mysteries, but some of
the extant inscriptions suggest that hymns were recited there. There are a few
short fragments surviving on the frescoes from the mithraeum at Santa Prisca
tentatively identified by Vermaseren as hymns48:

Nubila per ritum ducatis tempora cuncti

You must suffer through hard (cloudy) times together.
Dulcia sunt ficata avium sed cura gubernat
Sweet is the birds' liver, but care prevails.
Fecunda Tellus cuncta qua generat Pales
The land where Pales generates everything is fertile.
Accipe thuricremos, pater, accipe, sancte leones

46
For instance, on a few Mithraic reliefs we have a depiction of a sitting lion. Could it be a
sort of visual epithet of Mithras, a comparison with a lion?
47
For instance, on the relief from Neuenheim, CIMRM 1283.
48
M.J. Vermaseren, Mithras in der Römerzeit, in: Idem (ed.), Die orientalischen Religionen im
Römerreich, EPRO 93, Leiden 1981, pp. 96-120 (translations are mine).
14 Alexander Kirichenko

Holy father, receive the incense-burning lions.

Per quos thuradamos per quos consumimur ipsi
Through which we sacrifice the incense, through which we our-
selves are comsumed.
Fons concluse petris qui geminos aluisti nectare fratres
Spring concealed in rocks, you nourished the twins with nectar.
Hunc quem aur(eis) humeris portavit more iuvencum
This calf that he carried on his golden shoulders in the right way (?).
Et nos servasti eternali sanguine fuso
You saved us too by spilling the blood that makes us immortal.

There are a few dedicatory inscriptions that identify Mithras not only with the
Sun, but with a number of other gods as well, which is exactly what happens
in the Orphic (and other) hymns. Incidentally, the same inscriptions contain a
number of epithets applied to Mithras in the manner of the Orphic Hymns.

CIMRM 463. EÂw ZeÁw M¤traw ÜHliow kosmokrãtvr éne¤khtow. Di‹ Ñ

Hl¤ƒ megãlƒ Sarãpidi svt
ri ploutodÒt˙ §phkÒƒ eÈerg°t˙ éneikÆtƒ
M¤yrò xaristÆrion.
CIMRM 474. Di‹ ÑHl¤ƒ M¤yr&.
CIMRM 475. Di‹ ÑHl¤ƒ M¤yr& Fãnhti.

Even this meager evidence would suffice to suggest that prayers similar to the
Orphic Hymns might have been used in the mysteries of Mithras. However,
even if we did not have this evidence, just the fact that the cult of Mithras is
known to have appealed to the same kind of people as were active adherents
of other mystery cults where hymns similar to the Orphic ones were used
would indicate that this particular hymnal form was well familiar to the wor-
shippers of Mithras. The fact that one of the inscriptions cited above identifies
Mithras with the Orphic divinity Phanes might be significant as well (CIMRM
475).

Finally, there is one more piece of evidence that may corroborate my hypothe-
sis about the structural connections between hymns and Mithraic cult images,
even though I am truly at a loss as to how to interpret it: it is a small relief de-
picting Mithras with a torch (as Cautes?), at the bottom of which there is an
inscription saying Hymnus inbicto (CIMRM 646, CIL 3865), instead of invicto.
This relief does not have any side panels. What it might indicate is that, just as
 Hymnus invicto 15

a hymn, any Mithraic cult image was conceived of as a kind of timeless por-
trait of the god drawn in order to propitiate and praise him. It could be either
expanded by any number of episodes depicting Mithras’ beneficent activities
or qualities, or reduced to a bare minimum of the tauroctone scene or any
other representation of the god.

Alexander Kirichenko
Harvard Univesity
USA–Cambridge, MA 02138
e-mail: kirichen@fas.harvard.edu