University of Nebraska - Lincoln

DigitalCommons@University of Nebraska - Lincoln

Theses, Dissertations, and Student Creative Activity,
Art, Art History and Design, School of
School of Art, Art History and Design

4-2017

The Origins and Identity of Roman Mithraism

Charles R. Hill
University of Nebraska Lincoln, crhill8392@gmail.com

Follow this and additional works at: http://digitalcommons.unl.edu/artstudents

Part of the Ancient History, Greek and Roman through Late Antiquity Commons,
Ancient, Medieval, Renaissance and Baroque Art and Architecture Commons,
Classical Archaeology and Art History Commons, Classical Literature and Philology
Commons, History of Religion Commons, History of Religions of Eastern Origins
Commons, and the Indo-European Linguistics and Philology Commons

Hill, Charles R., "The Origins and Identity of Roman Mithraism" (2017). Theses, Dissertations, and Student
Creative Activity, School of Art, Art History and Design. 116.
http://digitalcommons.unl.edu/artstudents/116

This Article is brought to you for free and open access by the Art, Art History and Design, School of at DigitalCommons@University
of Nebraska - Lincoln. It has been accepted for inclusion in Theses, Dissertations, and Student Creative Activity, School of Art, Art
History and Design by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln.
THE ORIGINS AND IDENTITY OF ROMAN MITHRAISM

by

Charles R. Hill

A THESIS

Presented to the Faculty of The Graduate College at

the University of Nebraska In Partial Fulfillment of

Requirements For the Degree Master of Arts

Major: Art History

Under the Supervision of Professor Philip Sapirstein

Lincoln, Nebraska

April, 2017
 THE ORIGINS AND IDENTITY OF ROMAN MITHRAISM

Charles Hill, M.A.

University of Nebraska, 2017

Advisor: Philip Sapirstein

This thesis is a reassessment of scholarship concerning the origins of the cult

mysteries of Mithraism in its Roman form during the Imperial Period. While much has

been published in the debate over the cult’s true origins, we are still left without a

satisfactory answer. The present work is an attempt to reconcile some of the arguments

th th th st
posed in the 19 and early 20 centuries with those of the later 20 and 21 centuries,

focusing mostly on the cult’s art and iconography in Mithraea, the central spaces of

Mithraic worship. First will be a summary of scholarly opinion on the cult’s origins and

possible explanations for the cult’s later variations, followed by a section in which the

typical aspects of Mithraic spaces are established by region, to the extent that is possible.

Next will be a chapter in which specific sites in various regions of the Empire are

discussed in more detail, focusing on the dichotomy between the typical form of a

Mithraeum in that region and those aspects which point to variations between Mithraic

groups in different settings. Additionally, divergences in artistic representation between

spaces of civilian and military versions of the cult will be considered, as it is argued that

the distinction between these groups of worshippers is responsible for the development

of alternative aspects of the mysteries closer to the Roman core versus on the periphery

of the Empire. It is concluded that Mithraism, while consistent in many ways across the

Roman world and widely variable in others, was not exempt from the processes of

Roman religious syncretism, and is in fact one of the strongest examples attesting to its
efficacy. While this view is not new to the study of Mithraism, most recent scholars have

preferred to describe the cult as either relatively uniform across time and space, or as

entirely disjointed, to the extent that it should not be considered as a single cult tradition.
 iv

Table of Contents

2 Chapter One: Historiography

29 Chapter Two: Attributes of a Typical Mithraeum

45 Chapter Three: Survey of Specific Mithraea

64 Chapter Four: Assessment of Evidence

80 Figures and Illustrations

101 Works Cited

 v

Figures and Illustrations

80 Fig. 1: Typical tauroctony

th
81 Fig. 2: Perseus and Taurus constellations on 18 century star map

82 Fig. 3: Campbell Type I tauroctony

83 Fig. 4: Campbell Type II tauroctony

84 Fig. 5: Campbell Type III tauroctony

85 Fig. 6: Campbell Type IV tauroctony

86 Fig. 7: Campbell Type V tauroctony

87 Fig. 8: Campbell Type VI, Dura-Europos Zenobios relief

88 Fig. 9: Campbell tauroctony Subtype A

89 Fig. 10: Campbell tauroctony Subtype E, Osterburken relief

90 Fig. 11: Campbell Natural cave type relief

91 Fig. 12: Dura-Europos Mithraeum in situ

92 Fig. 13: Dura-Europos Ethpani relief

93 Fig. 14: Dura-Europos Magi paintings

94 Fig. 15: Dura-Europos hunting fresco

95 Fig. 16: Felicissimus Mithraeum mosaic floor illustration

96 Fig. 17: Basilica di San Clemente Mithraeum

97 Fig. 18: San Clemente altar

98 Fig. 19: Front of Dieburg Mithraeum relief panel

99 Fig. 20: Rear of Dieburg Mithraeum relief panel

100 Map 1: Mithraeum locations in the Roman Empire

 Introduction
The Roman mysteries of Mithras emerged around the time of the establishment of

the Roman Empire and persisted for several centuries, until their decline and

disappearance in the fourth century of the Common Era. The Mithraic cult, its

worshippers, and its activities have long been shrouded in mystery, as insider accounts of

the mysteries and cult practices simply do not exist. There is an abundance of

archaeological material extant today, however, which has been associated with the cult,

found across the Roman Imperial world, from as far east as modern Iran to as far west as

England and parts of Spain. Mithraic sites also dot the northern coast of Africa, and

appear along the Nile, standing opposite sites along Roman frontiers in Germany and

Eastern Europe. Despite the number of sites available for study, however, there has been

considerable disagreement from the beginning of modern Mithraic scholarship in the late

th
19 century to today as to the cult’s origins and its true identity. Did Mithraism descend

directly from the Indo-Iranian Avestan, Mazdean, and Zoroastrian traditions, or is the

connection between these and Roman Mithraism limited to the name of the central deity

himself? To what extent was the cult distinctly Roman, and to what extent did it

permeate or defer to local traditions as it expanded across the Imperial world? In this

thesis, I seek not necessarily to provide new answers to all of these questions, but to

mediate between schools of thought which are highly polarized on the issue of the cult’s

origins and unique identity as a part of the Roman religious tradition. I will begin with a

historiography in which I detail the most prominent arguments in the debate over time.

Following this will be a section in which I establish those characteristics most typical of

the cult’s iconography and artistic program in different regions. I will then describe in

greater detail certain sites within those regions, focusing both on the commonalities
 1

between sites and on those things which render each site unique. Finally, I will conclude

with an evaluation of the evidence from these particular sites, in an effort to demonstrate

that Roman Mithraism is both exceptional and unique among religions adopted by the

Romans, and also one of the single best examples of the syncretistic processes that

defined Roman religious tradition throughout history. I will demonstrate that, while

Mithraism deserves its designation as a mystery cult in every possible sense of the

phrase, there is still much we can say about the religion, and an immense deal that we

might learn from its study.

 2

Chapter One
Historiography
Mithraism, the adopted cult of the Roman Empire, often evokes ideas of

mysterious and shadowy worship of a violent and powerful Eastern god by groups of men

th
in subterranean chambers, hunched over offerings of food and drink. Throughout the 20

century, scholarship on Mithraism has mostly perpetuated this vision of the cult.

Mithraism was originally assumed to be a highly exotic religion with an origin far outside

the Roman Imperial world both chronologically and geographically, with few scholars

arguing against the earliest opinions that the cult was always primarily Indo-Iranian, and

somewhat antithetical to the values of Roman civilization. However more recently some

debates have also emerged concerning the single-sex exclusive nature of the religion,

with some scholars arguing that the depiction of female deities alongside Mithras and

other Zoroastrian figures in Mithraea, along with the recording of female benefactors of

sanctuaries, are evidence of mixed-gender worship. Alongside this re-examination of

Mithraism as a male-exclusive cult has arisen scholarly skepticism about the rejection of

all aspects of the feminine by Mithraic worshippers. Given the supposed Eastern origin of

Mithraism and the god Mithras himself, it would be highly unlikely that Roman men

never considered the ramifications of worshipping a deity who dressed in a clearly

Eastern (and therefore inherently feminine) fashion. Indeed, the Eastern and the feminine

often seem to have been inseparable in the minds of ancient Mediterranean peoples such

as the Greeks and then later the Romans. Therefore, it is necessary to consider the

treatment of the feminine principle in establishing the degree of “Roman-ness” of any

adopted religion. In this historiography, I seek to provide a concise summary of the ways
 3

in which a few key scholars have addressed the debate over the Eastern origins of

Mithraism in the Roman world, and to situate my own ideas within the context of that

particular debate. To begin, I will address some of the most often-cited ancient written

sources which mention the cult, although I will temper this with cautions about the

validity of those writers in particular, given their interactions with the cult. Then I will

th
summarize the 19 century and early 20th century scholarship on Mithraism which

established it as Indo-Iranian in origin from the start, and later examine some of the bold

counter arguments which emerged nearly 70 years later that sought to scrap all notion of

Roman Mithraism’s actual ties to Indo-Iranian religions. This debate represents an

interesting intersection between the disciplines, and requires some thought concerning the

extent to which art represents or belies the circumstances of the real world.

Primary sources for any ancient mystery cult can be difficult to evaluate, but this is

especially true for the case of Roman Mithraism. First and foremost there is the issue of the

name Mithras, which appears in myriad forms across a wide array of languages hundreds of

years before Roman Mithraism came about. Fortunately early Mithraic scholars went to great

lengths to track down whatever mentions of Mithras or his worshippers could be found in

1
texts of ancient Europe and Asia Minor. However any reader of the ancient sources should

be cautioned against assuming that the appearance of the name Mithras is sufficient evidence

to argue that the source in question is concerned with what we now call Roman Mithraism.

Especially in sources from the first centuries

1 Alfred Geden, Select Passages Illustrating Mithraism. MacMillan, 1925. Geden’s work is drawn from an
earlier Cumont volume in which untranslated passages were compiled and commentaries were given in French.
Geden provides English translations of the original passages and provides some of his own commentary. Geden’s
work is an excellent starting point for the interested reader, but Cumont’s 1896 Textes et Monuments Figures
Relatifs aux Mysteres de Mithra provides a fuller list of ancient mentions, for those who can read French.
 4

BCE and CE it can be exceedingly difficult to distinguish between the Mithras of the

Roman mysteries and Mithras of the Avestan and Mazdean traditions. In fact it was likely

th
thanks to such confusion that the earliest modern Mithraic scholars of the 19 century so

readily accepted a strong direct link between Mithras in Indo-Iranian traditions and

Mithras in the Roman mystery cult.

Sometime in the first century CE, Plutarch described Mithras as a mediator

between the two opposing forces of light (named Horomazes) and darkness (named

Areimanius) in Persian religion, following the traditions set forth by Zoroaster 5000 years

2
before the Trojan War. It is fairly clear at this point that Plutarch is still describing

Mithras within the context of traditions foreign to the Roman religious program, as he

names Horomazes (sometimes Ahura-Mazda) and Areimanius (sometimes Ahriman), two

3
figures who were likely not carried over into the Roman mysteries to any large extent.

Plutarch also mentions Mithras by name in a few other works, but most of these

appearances are passing mentions or invocations of the god by name only. In De Fluviis,

however, we find the story of Mithras impregnating a living rock in order to produce a

son, and it is supposed that the god did so out of his contempt for women and the

4
feminine. While this account of the god would certainly fit with the notion that only men

worshipped him in the Roman world and shunned the influence of the feminine, it is also

difficult to reconcile with imagery from Roman Mithraic spaces, as more often it is

Mithras himself who appears being born from the living rock rather than impregnating it

2 Plutarch, De Iside et Osiride 46. trans. Geden 28-30.

3 This negatively anticipates the arguments set forth by Franz Cumont slightly, but the prevailing scholarly
opinion today is that references to Ahura-Mazda and Ahriman within the Roman version of the cult are simply not to
be found. Geden also notes in his commentary on this passage that Plutarch’s sources for this description of Persian
traditions are hazy at best.
4 Plutarch, De Fluviis 23. trans. Geden 30.
 5

to create a son. Thus this may be an aspect of earlier Persian tradition that was adopted by

Roman Mithraists and altered slightly to fit with Roman Mithras’ identity as a

miraculously born savior deity.

More significant than early and unclear mentions of Mithras are those found in

Porphyry and Tertullian’s writing. These are the authors most often cited in modern

debates over the identity and practices of Roman Mithraism, and therefore the most

important to evaluate and understand. I will return to some of the passages mentioned

here in my description of the debate over women and their participation in the Roman

version of the Mithraic mysteries, as they form the ancient literary basis for much of the

argument.

Tertullian, writing in the late second and early third centuries, described a number of

the practices of Mithraic worshippers, stating that they would be proffered a crown during an

initiation ceremony, and expected to reject the crown in a symbolic gesture, showing that the

5
only crown they accepted was the presence of Mithras himself. It is also implied that this

practice was meant to be a mockery of Christian martyrdom, and Tertullian certainly implied

elsewhere that certain Mithraic rites were shams of Christian sacraments, stating that the

devil used these to deceive the worshippers of Mithras, as he “marks his own soldiers with

the sign of Mithra on their foreheads, commemorates an offering of bread, introduces a mock

6
resurrection, and with the sword opens the way to the crown.” Thus it is clear that Tertullian

was decidedly not interested in the truth of the Mithraic rituals, and rather more in

denigrating their rites and beliefs as bastardizations of Christianity. It is therefore difficult to

trust anything he claims to know about the

5 Tertullian, De Corona 15. trans. Geden 43.

6 Tertullian, De Prescriptione Hereticorum 40. trans. Geden 44.
 6

mysteries. However another, much shorter passage appears later in this paper as a

justification for the tentative identification of a site in North Africa as Mithraic when

there is little evidence to confirm that. A short passage in which Tertullian stated that,

7
“The lions of Mithra are represented as types of an eager and impetuous nature.” has

been used, controversially, to identify sites as Mithraic based on the presence of lion

8
imagery, even in the absence of any other evidence. Overall, then, it is difficult to put

much confidence in Tertullian as a viable source for information on the values or

practices of Roman Mithraists.

Porphyry, on the other hand, seems to have written much more objectively about

Mithraism, spending much more time establishing the historical context for their rites and

beliefs than on denouncing their activities. In De Antro Nympharum, Porphyry connects

the worship of Mithras to the consecrations made by Zoroaster, although unlike

Plutarch’s Zoroaster, Porphyry insists that Mithras was the central deity in this new

9
religion, rather than the mediator between light and dark that Plutarch described. In

addition to this, Porphyry also claims that Zoroaster was responsible for the Mithraic

10
habit of worshipping in natural or constructed cave-like spaces. However, more

importantly, it is in Porphyry that we find one of the most controversial passages about

Mithraic practice with his claim that, “the mystics who take part in the actual rites are

called lions, the women hyenas, the servants crows, and of the fathers… for these bear

11
the names of eagles and hawks.” This passage has been debated hotly in more recent

7 Tertullian, Adversus Marcionem. 1.13. trans. Geden 43.

8 It is to this passage that David refers later in arguing that the site in Oea, North Africa is Mithraic.
9 Porphyry, De Antro Nympharum 5-6. trans. Geden 47-48.
10 Ibid.
11 Porphyry, De Abstinentia 4.16. trans. Geden 50.
 7

years, as it appears to provide evidence for women’s participation in the actual Mithraic

rites. It has been pointed out that the text is corrupted near the word “hyenas,” and that it

12
may also be the word “lionesses.” However it is not clear that either of these

translations of the word would point confidently to women’s involvement in the cult.

Therefore while Porphyry is less hostile to Mithraism than Tertullian, it is not clear that

he understood the cult overly well either. The vast majority of other ancient sources

mentioning Mithras or the Roman mysteries followed Tertullian’s lead, as many of them

were also Christians and therefore hostile to the cult. Thus we see that, in the absence of

insider accounts of the cult’s traditions and practices, ancient texts on the mysteries are

13
largely unsatisfactory, and require a healthy amount of skepticism in their treatment.

Some of the earliest modern published work on Mithraism and its associated

th
archaeological remains and imagery came about near the end of the 19 century thanks

to Franz Cumont, a Belgian archaeologist and historian. Cumont visited and participated

in the excavations of a great number of Mithraea, and synthesized many of his notions

concerning the cult based on his observations of single sites and their interactions with

one another. In particular, Cumont argued in a matter-of-fact way that Mithraism in the

Mediterranean was inextricably linked to the Mazdean, Avestan, and Zoroastrian

traditions of the Persians and their ancestors. Additionally, he was one of the first

scholars to argue for the gender-exclusivity of the cult based on his observations of

iconography and inscriptions related to the cult’s spaces and monuments. While many of

12 See notes 46 & 50 for more on the specifics of this debate.

13 In addition to those discussed here, see the following later authors: Justin Martyr, Apology 1.66; Dial. c.
Tryph. 70 and 78.
Firmicus Maternus, De Errore Profanarum Religionum chapters 4 and 20.
Jerome, Ep. CVII. ad Laetam; Adv. Jovinianum 1.7 and 2.14; Comm. in Amos, 5.9-10.
 8

Cumont’s ideas have persisted in more recent scholarship on Mithraism, E.D. Francis

notes in a preface to his translation of one of Cumont’s articles that, “In his quest for

coherent synthesis, however, Cumont sometimes pressed his conclusions beyond the

available evidence, and what many epigone have on occasion taken to represent an

unassailable judgement may rest on little more than an imaginative interpretation of

14
unusually problematic data.” This is unsurprising, given that Cumont was working

almost entirely from archaeological, non-textual evidence, with little to no other prior

scholarship to cite. In reality, it may not have been Cumont’s data which was

problematic, but his use and interpretation of that data in an attempt to answer questions

which it simply could not be stretched to answer. Additionally, a dearth of textual

primary sources for Mithraism and the activities of its worshippers meant that Cumont

was working from an already incomplete, purely archaeological/visual record in order

to answer questions about intangible aspects of the religion, including issues of gender,

foreignness, and interaction with other adopted cults.

th
It is evident that even by the middle of the 20 century, this early Mithraic

scholarship, along with Mithraic art in general, was due for a re-evaluation, taking into

account numerous discoveries made in the intervening years. Despite the acknowledged

faults in Cumont’s research, however, Francis also notes that, “Although his essay is

necessarily a work as much of synthesis as of original research, Cumont provides a valuable

commentary on the iconography of the decoration and the liturgical implications of the

15
graffiti.” In this commentary of iconography on the Dura Mithraeum, Cumont

14 E. D. Francis, Editor’s Preface to Franz Cumont, “The Dura Mithraeum.” in Mithraic Studies. ed. John
R. Hinnels: Manchester, 1975. 151.
15 Ibid., 153.
 9

states quite clearly that, “it is certain that members of the female sex were never admitted

into the service of the soldiers’ god. In no Mithraeum do we find any mention of a

16
woman serving as priestess, initiate or even as a donor.” Having stated this, Cumont

makes no other mention of the feminine or female involvement in the cult of Mithras.

This is unsurprising, as gender studies related to most religions of the ancient world were

th
hardly in the forefront of scholarly concern until later in the 20 century, and certainly

not when it came to Mithraism, most widely accepted to be male exclusive. Whereas a

re-evaluation of evidence for gender inclusivity of other supposedly single-sex (likely

women-only) cults may have been justified by the primarily male body of scholars,

Cumont simply did not write in such a way that there seemed to be any need to question

his conception of male-centric Mithraism. In this way, early scholarship on Roman

adopted religions seems to have hewn fairly close to Roman senatorial opinions on

mystery religions; secretive cults only open to women needed to be investigated and

brought into the light, but there was no need to treat secretive all-male cults in the same

way.

Indeed, Cumont did not concern himself much at all with the gender of Mithraic

worshippers, although his ideas set the precedent for much later scholarly assumption that

all Roman Mithraic adherents were men. In much the same way, Cumont’s ideas

concerning the origins of Mithras himself and the many symbols found extant in

Mithraea around the Roman world paved the way for many later Mithraic researchers up

until the 1970s to presume that Roman Mithraism had strong ties to Indo-Iranian and

Zoroastrian traditions. As David Ulansey noted in 1989, it was not until the First

16 Franz Cumont, “The Dura Mithraeum.” Ed. and trans. E.D. Francis. Mithraic Studies. ed. John R.
Hinnels: Manchester, 1975. 199.
 10

International Congress of Mithraic Studies in 1971 that Cumont’s ideas faced radical

th
rebuttal, though he had had to defend some of his more tenuous claims earlier in the 20

17
century from skeptical researchers. Ulansey states that, “From that moment on, it could no

18
longer be assumed that Roman Mithraism originated in Iran.” Thus it is apparent that 1971

was one of the major turning points in the history of Mithraic scholarship, and the first time

Cumont’s venerable body of work was re-evaluated wholesale.

Before this rejection of Cumont’s ideas, however, several other scholars produced

impressive volumes following his template for the study of Mithraic origins. Also

interested in the connections between Roman Mithraism and Zoroastrian traditions,

Leroy A. Campbell, a religious historian publishing in the 1960s and 70s, discussed

much of the iconography of various scenes found in Mithraic structures. In his 1968

book, Campbell examines at great length the various aspects of the god Mithras, and the

relation of those aspects to various divine figures in Mediterranean and near-Eastern

religious traditions. Campbell classified various pictorial representations of aspects of the

god Mithras, including different types of the tauroctony scene in which Mithras appeared

in the act of slaying a bull, often accompanied by a number of animal familiars who

engage with him and with the sacrificial bull in various ways (see Fig. 1 for example).

Not only was Campbell’s work heavily invested in language and Mithraic relations to

predecessor Indo-European traditions, but it was also heavily reliant upon images of

Mithraic scenes as well. Despite some of the criticisms already leveled against Cumont’s

readiness to connect Roman Mithraism to earlier Indo-European divinities, Campbell had

17 David Ulansey, The Origins of the Mithraic Mysteries. Oxford, 1989. 9-10.
18 Ibid., 12. Ulansey also notes here that no alternative theories were immediately proposed to explain the
origins of Roman Mithraism, leaving somewhat of an intellectual vacuum in a field which had until then been
relatively stable.
 11

the advantage not only of having Cumont’s work to draw upon, but several decades’

worth of additional discoveries concerning Mithraic imagery following Cumont’s death

in order to argue his case more soundly. It is important to include Campbell’s work in

any discussion of the development of Mithraic scholarship, as his efforts to prop up

Cumont’s idea of a Mithraic-Zoroastrian connection came about just prior to the

conference at which many of Cumont’s notions came under attack for having been based

on evidence too circumstantial to take seriously. For both Cumont and Campbell, then, it

was vital to prove a strong connection between Roman Mithraism and Zoroastrianism,

likely in an effort to explain some of the exotic aspects of the cult, such as Mithras’

Phrygian accoutrements, the appearance of a number of different helper figures in

various tauroctony scenes, and the depiction of other Eastern (or simply non-Roman)

figures alongside Mithras in some Mithraea.

Campbell, while not primarily concerned with aspects of the feminine in

Mithraism (and certainly not concerned with the actual gender or sex of Mithraic

worshippers), does address the symbolic interaction of the feminine and the masculine in

tauroctony scenes. Essentially, Campbell promotes the idea that the snake helper which

appears in most tauroctony scenes might be a representation of the feminine, interested in

capturing some of the generative powers from the blood of the bull, a bi-sexual entity,

19
once they have been released by the actions of the masculine figure, Mithras himself.

Rather than describing the interaction between the binaries of the masculine and

feminine, Campbell was simply interested in the active figures of the scene, regardless of

their gender associations, though it is inevitable that the masculine gets more attention,

19 Leroy Campbell, Mithraic Iconography and Ideology. Brill, 1968. 21 and 248-9.
 12

simply because it is more prevalent in a tauroctony. Indeed, Campbell seems to have

been drawn in by Cumont’s notion of dualism between good and evil represented by the

various helper figures present in tauroctony scenes, although Ulansey notes that these

were ideas which soon came under fire during the 1971 break from Cumont’s traditional

20
arguments. Campbell’s volume is vast and extraordinarily rich, including discussions

not only of Roman Mithraic sites, iconography, and inscriptions, but also of Indo-Iranian,

Avestan, Mazdean, and Zoroastrian sites and passages in which he saw likely sources of

inspiration for the later Western Mithraic traditions.

It would be impossible to consider all of Campbell’s evidence within the scope of

a thesis such as mine, but it is mostly to Campbell’s work here that I will refer in

establishing the norms of Roman Mithraic sites (to the extent that this is possible), simply

because this volume is one of the most comprehensive collections of evidence available

to a modern Mithraic scholar, regardless of one’s opinions on his connection between

Mithraism and Eastern traditions. It is also important to acknowledge Campbell’s vital

recognition, somewhat more advanced than Cumont’s, that Mithraism was by no means a

uniform tradition, either in the Roman world or in its various manifestations throughout

21
the rest of the Mediterranean and the Middle East. While this may seem obvious, given

the breadth of evidence provided in Campbell’s volume, not all scholars working in the

th st
later 20 and early 21 centuries acknowledge this properly, leading to a somewhat

distorted picture of Mithraic practice in the Roman Imperial world. While Campbell

certainly espouses the same ideas as Cumont in regards to Western Mithraism’s origins in

20 Ulansey, Origins, 10-11.

21 Campbell, Mithraic Iconography, 4.
 13

the East, his work is a careful consideration of the variability in Mithraic sites of the

Roman Imperial world nonetheless.

Writing in response to Cumont’s early works, in which he established many of

his most fundamental theories concerning Mithraism, R. L. Gordon, a social/cultural

historian, argued in the 1970s that Cumont’s works represented an attempt to force

Roman Mithraism to fit with incompatible doctrine from its Zoroastrian origins. But he

also cautioned against scrapping Cumont’s ideas entirely, stating that, “Criticism of

Cumont is only a preliminary to the elaboration of an alternative view of Western

Mithraism which is content to accept the fact that we do not possess a great deal of the

information about the cult which was positively demanded by Cumont’s conception of

22
his task.” This being said, Gordon then addressed problems in Cumont’s assumption of

Roman Mithraic direct doctrinal descent from Zoroastrian origins, but does not exactly

offer satisfactory alternative explanations for the origins of Mithraism, leaving the

connection between Indo-Iranian Mithras and Roman Mithras somewhat unclear in the

wake of his dismissal of Cumont’s ideas. As noted by Ulansey, Gordon was one of the

first scholars, along with John Hinnells, to abandon the Cumontian template and begin to

look very critically at the evidence hitherto presented for Roman Mithraism’s strong ties

to Indo-Iranian traditions.

Gordon revisited Mithraism frequently in later publications, and in 2000 he

published a translation of The Roman Cult of Mithras: The God and His Mysteries,

originally published in 1990 by the German historian Manfred Clauss. Focused not only on

images, but on archaeological remains as well, Clauss wrote what is now widely

22 R. L. Gordon, “Cumont and the doctrines of Mithraism.” Mithraic Studies. Ed. John R. Hinnels:
Manchester, 1975. 220-1.
 14

considered to be one of the best and clearest overviews of Mithraism as a whole in the

Roman world. Perhaps most importantly, Clauss is aware of the variable nature of cult

worship in the Roman Imperial world, and cautions his reader early in the text, stating, “it is

unsatisfactory to speak of the mysteries of Mithras as a unified religion. To do so makes

23
things simpler, but it also gives a false impression.” Indeed it is important to recognize this,

especially when considering arguments concerning the earliest origins and forms of Roman

Mithraism. In addition to getting at the source of the Mithraic mysteries, the existence of

varied Mithraic traditions and practices would have had other effects on worshippers.

Turning to one of the other more persistent debates within Mithraic scholarship, even if

regional variations of the cult allowed for the inclusion of women and/or feminine imagery in

Mithraea, it would be a great stretch to promote those deviations as evidence of gender

inclusion in Mithraism more broadly. And if it is possible that something long thought to be

one of the core tenets of Mithraic practice was not universally true, then it would certainly be

naïve to dismiss any lingering Eastern Mithraic practices within the Roman world as pure

invention or misinterpretation of evidence. But looking at Roman Mithraism more generally,

Clauss stated that, “no direct continuity, either of a general kind or in specific details, can be

demonstrated between the Perso-Hellenistic worship of Mitra and the Roman mysteries of

24
Mithras.” This sounds, then, like a final rejection of Cumont and Campbell’s ideas of

Roman Mithraism’s direct descent from Iranian traditions, and for the most part recent

scholarship has viewed this debate as settled. As a result, much of Mithraic scholarship has

now become concerned

23 Manfred Clauss, The Roman Cult of Mithras: The God and His Mysteries. Trans. Richard Gordon:
Routledge, 2001. 16.
24 Ibid., 7.
 15

more with the idea of gender-exclusivity in the Roman tradition of Mithraism alone. If

the Roman version of the cult is truly distinct from any previous Eastern traditions, are its

attitudes towards the feminine also unique? In many other cases, Roman society proved

to be overall hostile to feminine or anti-masculine aspects of other adopted cults. Was

this the case with Mithraism? Or did the presence of such an aggressively masculine

central figure such as Mithras allow for the subtle inclusion of the feminine symbols in a

tradition which was otherwise exclusively male? These are questions I have considered

in past writing, and ones which cannot fully be answered until it is understood more

clearly whether Roman Mithraism is indeed distinctly Roman, or descended more

directly from Indo-European traditions. Not only, then, are the early ideas of Cumont at

stake when we consider Mithraism’s origins, but many other questions about Mithraism

necessitate a more confident answer as to the identity of Western Mithraic worshippers,

and their understanding of their own cult’s origins.

However, there have been some other recent efforts to explain the origins and

identity of Roman Mithraism by scholars whose stated goals were to reevaluate Western

Mithraism through the examination and synthesis of evidence ignored or overlooked in

many other texts on the cult. In 1980, Michael Speidel broke from the tradition of trying

to reconcile Mithraism with any Indo-Iranian predecessor traditions, and instead made an

argument that Roman Mithraism came about as descendant of a Greek cosmological

religion. He rejected Cumont’s ideas that Roman Mithraism retained strong ties to Iranian

traditions, saying, “The opposite is true. Mithraism is originally and substantially a Greek

25
religion with only a few Iranian elements.” The author continues on to claim that a

25 Michael P. Speidel, Mithras-Orion: Greek Hero and Roman Army God. Brill, 1980. 2.
 16

number of previously overlooked ancient sources can be reconciled neatly with Mithraic

imagery and art to demonstrate a connection between the constellations surrounding that of

26
Orion and the story of Mithras. Speidel accurately points out that the central tauroctony

scene is the most persistent indicator of Mithraic activity at any site, but claims that the

fields which have traditionally been employed to explain the scene have proven incapable of

generating a comprehensive narrative. The author rejects the idea that Iranian religious

literature, Greek and Roman art, and astral and seasonal symbolism could ever fully explain

the mysteries of the tauroctony, and suggests the inclusion of Greek and Roman astronomy

in Mithraic scholars’ toolkits in order to decipher the true meaning of the bull-slaying

27
iconography. Importantly, Speidel notes that his is not the first attempt to expound upon

the celestial origins of the mysteries, acknowledging the very early attempt by K.B. Stark in

28
1868 to do the same. However Stark’s ideas were quickly swept aside by Cumont’s, and

largely forgotten for a number of decades.

Speidel cites the written work of Porphyry at length in arguing for Mithras’

connection with Orion, although he is somewhat vague on certain details, leaving some

doubt as to whether the passages cited pertain to Mithras or Orion originally. The author

does, however, propose a tidy explanation for the absence of any textual connection

between the names of Mithras and Orion in the ancient world, speculating that, “The

reason may be that within the cult to call Mithras Orion would have meant to deprive

the god of his true name, while outside the cult there was a certain reluctance to reveal

29
religious secrets.” While this is certainly not beyond the realm of possibility,

26 Ibid., 3.
27 Ibid., 5.
28 Ibid., 6.
29 Ibid., 26.
 17

considering the Roman practice of syncretizing various adopted gods together with their

own native deities, the issue is somewhat confused by something else the author

mentions, as he asserts that depictions of the hero Orion slaying the bull with a sword

30
are more typical of the Roman period than the Greek. Therefore it would be difficult to

argue for a truly Greek origin of the Orion-Mithras cult, if it were only a Roman

reimagining of the hero Orion to which Mithras would later be connected. However

Speidel pushes the Orion connection in the origins of Mithraic imagery, arguing that the

snake and scorpion often seen in the tauroctony, lapping at the bull’s bodily fluids as it

is slain, are not Mazdaean figures representing evil, but are neutral or benevolent figures

31
drawn from Greek constellations. Therefore this author is contending that the Roman

practice of syncretization is relatively simple, and not capable of drawing influence from

more than one or two different cults in order to form a new one. Overall this seems far

too simplified, and I will argue later that it is far more likely that the predecessor

traditions contributing to what we now label Roman Mithraism were many.

Similarly to Speidel, David Ulansey argues for a cosmological origin of the stories

and images of Mithraism, although he connects Mithras not with the great hunter Orion, but

with the Greek hero Perseus. In the introductory section of his monograph, Ulansey notes

appropriately that the most advantageous thing for modern Mithraic scholars is the

32
persistence of the tauroctony at Mithraic sites. However Ulansey also preliminarily

dismisses the Cumontian interpretation of some of these symbols such as the dog and snake

as evidence of the Iranian dualistic struggle between agents of good

30 Ibid., 25.
31 Ibid., 29.
32 Ulansey, Origins, 6.
 18

and evil, pointing out that scholars have long argued for a certain amount of indifference

33
in the attitudes of these figures. Like Speidel, Ulansey also refers to Stark’s early ideas

concerning Mithraism’s connection to the stars, though he does not see as much

34
incompatibility between these ideas and Cumont’s as Speidel suggested. It is important

to note here that Ulansey carries on to reject directly Speidel’s connection between Orion

and Mithras, as he points out that not all the equatorial constellations are represented in

the tauroctony, and there is therefore no reason to force a connection between Orion and

Mithras simply because of his proximity to other constellations appearing in the scenes of

35
Mithras fighting and slaying the bull. Rather, Ulansey argues there is a much more

sensible template for Mithras already evident in the sky near the Taurus constellation,

bearing many of the symbolic attributes of Mithras, perfectly ready to be syncretized.

The author points out that the constellation of Perseus, wearing a Phrygian cap and

holding a dagger, is a much tidier parallel to Mithras than Orion, not only because of his

accoutrements, but also as Perseus appears in the night sky above Taurus (Fig. 2),

36
whereas Orion is problematically separate from the bull. Ulansey even pushes the

connection further, arguing that the mythological connection of the hero Perseus and the

nation of Persia is further proof of compatibility with Mithras, who is almost always

37
depicted in Oriental garb. Additionally, the author even claims that a connection with

the hero Perseus might offer some explanation of the typical portrayal of Mithras looking

away from the bull as he kills it, as he sees a parallel with Perseus averting his eyes from

33 Ibid., 10-11.
34 Ibid., 15-16.
35 Ibid., 22-23.
36 Ibid., 26-27.
37 Ibid., 29.
 19

38
the Gorgon Medusa as he fights her. Finally, and more importantly for my own

purposes, Ulansey identifies another piece of evidence for the Mithras-Perseus

connection. Citing Plutarch’s assertion that Mithraism originated with the pirates of

Cilicia in the first century BCE, Ulansey reinforces the idea that Perseus, the

mythological founder of Tarsus, would naturally appear in later cult as the adapted and

39
transformed deity Mithras. The only explanation Ulansey offers as to the process by

which Perseus would then have been transformed into Mithras and given the Iranian

name is somewhat unsatisfying, as it involves a tenuous connection between King

Mithridates VI of Pontus and his namesake. While it is true that Mithridates translates

40
literally as “given by Mithras,” it seems quite a leap to say that the pirates would have

taken Mithridates’ namesake as the new name for Perseus, so many generations

removed from the first appearance of the name Mithridates. Of course, this does not

disqualify the argument entirely, but it would be too bold to assert that this was the only

factor in the creation of the name Mithras for the god of the Cilician pirates.

Whereas many of the scholars I have discussed up to now focused primarily on

iconography and archaeological remains in investigating the origins and spread of Roman

Mithraism, Roger Beck sought in a 1998 article to reexamine the question of who would

have been responsible for the cult’s proliferation in the first and second centuries CE.

Beck responsibly stresses the hypothetical nature of his creation of a founding group for

the cult, but notes the importance of the questions which might be raised and answered in

38 Ibid., 31. Ulansey also continues on the next few pages to press this connection, arguing that there is a
remarkable similarity between certain depictions of Gorgons and those of the Lion-headed figure often found in
Mithraea.
39 Ibid., 43.
40 Ibid., 89-90.
 20

attempting to deduce the identity of the first adherents of a recognizably Roman version

41
of Mithras. The author summarizes and acknowledges many of the theoretical origins

of Mithraism which I have described above, but states that his own explanations are most

42
easily reconciled with some of Cumont’s ideas on Mithraic origins in Anatolia. In fact,

one of Beck’s criteria in identifying the earliest adherents of Mithras in the Roman world

43
is a familiarity with both Iranian religious traditions and Western astrological studies.

Not only does this fit nicely with Cumont’s insistence on a strong Indo-Iranian vein

within Roman Mithraism, but it also leaves considerable room to argue that Ulansey’s

idea of an origin amongst the Cilician pirates is tenable. Some difficulties remain,

however, as Beck points out the problem of apparently contemporaneous development of

the cult in fairly far-flung corners of the empire, usurping some earlier scholarly notions

that the cult only reached as far east as Dura-Europos after developing in the Western

44
parts of the Roman world. Finally, Beck reveals his candidate for the founders of

Roman Mithraism as the soldiers and civilians of the Commagenian dynasty in the first

century CE, claiming that this would explain its quick transmission, as Commagenian

troops would have had significant contact with members of the Roman military during

45
this period. This account of the cult’s origins is particularly interesting as it does not

entirely reject the camps of any of the scholars I have mentioned previously in this

historiography, but instead seeks to reconcile and update many of the strongest

th th
arguments from the 19 and 20 centuries.

41 Roger Beck, “The Mysteries of Mithras: A New Account of Their Genesis.” The Journal of Roman
Studies 88 (1998): 117.
42 Ibid., 116.
43 Ibid., 119.
44 Ibid., 118.
45 Ibid., 121-122.
 21

It would be foolish, in a summary of Mithraic scholarship, not to mention the

other popular debate related to the Roman mysteries of Mithras. The idea that the cult

was exclusively open to male worshippers has been around from the very beginning of

modern Mithraic research, and has largely remained unchallenged. While many early

Mithraic scholars were not particularly concerned with the gender-exclusivity of Mithraic

worship, they all laid substantial groundwork for later scholars interested in the conflict

between the masculine and the feminine in all aspects of Roman Mithraism. Some

scholars argued that an exclusively male cult would have demonized or otherwise

shunned all aspects of the feminine in order to bolster male worshippers’ masculine self-

assurance, but I think this would be naïve at best. Given the Roman capacity for

collective memory and deference to the mos maiorum, it would be highly unlikely that a

Roman male would not immediately recognize the often Phrygian dress of the god

Mithras as something Eastern, and therefore something less than completely masculine. It

would also be difficult to argue with any degree of success that any cult of the Roman

imperial world could successfully exclude all aspects of either the masculine or feminine.

While Mithras himself was sometimes shown to have literally been born of the living

rock, this does not itself preclude feminine involvement in his origin, as the Romans,

borrowing from the Greek tradition, would likely have seen the earth or living rock as a

feminine entity. Additionally, the bull, though typically associated with male deities in

Greco-Roman tradition, would likely have been seen as a representation of the feminine,

or at least of the not-masculine. Given the tradition of portraying male gods triumphing

over wild beasts as a representation of West triumphing over East, male triumphing over

female, and civilization triumphing over nature, the bull would certainly have been seen
 22

to stand for all of these things to the average Roman viewer. While settling the debate

over gender-exclusivity and tensions in Western Mithraism does not bear directly on my

endeavor to identify the origins and foundational practices of the cult, questions of East

vs. West in the Greco-Roman Mediterranean will always be entangled with issues of

masculinity and femininity. Therefore, gender and its treatment in cult iconography will

help to answer some questions as to the cult’s origins, and vice-versa.

The strongest voices in the debate over gender-exclusivity in Roman Mithraism

st
have come to the fore only in the 21 century, most notably in the debate between

Jonathan David and Alison Griffith. David, writing in 2000, seems to believe his is the

first serious attempt to argue against the entrenched position of all previous Mithraic

scholarship, that women were indeed excluded from the mysteries of Mithras.

Unfortunately, he seems to have fallen into the same trap as some of the first Mithraic

scholars, such as Cumont, in that he stretched a small amount of unsubstantiated evidence

far past the point of relevance. However, in light of the boldness of David’s claim, it is

vital to focus at greatest length in this historiography on his arguments, and the

subsequent categorical response to those arguments published by Griffith in 2006. I will

first lay out David’s arguments, and then report Griffith’s response to each of his major

points in the following section.

th th
In his article from 2000, David sets out to overthrow the 19 and 20 century

notion that Mithraism excluded women entirely, promising to draw upon epigraphic,

iconographic, and textual primary source evidence to demonstrate that, “the theory of

46
universal female exclusion from Mithraism is untenable.” While this claim would fit

46 Jonathan David, “The Exclusion of Women in the Mithraic Mysteries: Ancient or Modern?” Numen 47
(2000): 121.
 23

comfortably within the traditions of Mithraic scholarship if framed as Clauss recommends

Mithraism be viewed (as a sum of many diverse local Mithraic traditions rather than a rigid

and uniform tradition to which the congregations of all Mithraea adhered), David does not

make it abundantly clear that his intent is to argue for the inclusion of women as a local

variation, and instead sets out as if to suggest that women were included in the rites

everywhere. As his foundational piece of evidence, David cites a passage from Porphyry in

which animal names are given to the different grades of Mithraic initiation and participation.

Despite the possibility that the passage in question is both corrupt and inconsistent in various

manuscripts, David argues that Porphyry’s designation of women as either hyenas or

lionesses refers specifically to women involved with Mithraism, even going so far as to say

47
that Gordon’s rejection of this notion is inconsequential. David’s then turns to

archaeological evidence with the description of a pair of sarcophagi from Oea, North Africa,

which he argues are rife with Mithraic imagery similar to that found in a Mithraeum in

48
Rome, where a figurine of a woman also appears. The author continues on to reference

accounts of Mithraic rites given by Tertullian, arguing that any amount of inconsistency with

the established Mithraic tradition found outside of North Africa is reason enough to give

49
credibility to Tertullian’s otherwise accounts. In the following section, David proposes a

reevaluation of previous scholarly identification of the animals associated with the Mithraic

grades of initiation, asserting that in some cases the significance of those animals has been

misconstrued. Finally, despite his bold statements earlier in the article, David summarizes his

arguments

47 Ibid., 124-125.
48
Ibid., 125-126.
49 Ibid., 126-127. David notes that Tertullian is mostly concerned with North Africa in his writings, and sees this
as suggestive of a distinct regional sect of Mithraism in which women were involved as members.
 24

and concludes with a recognition that the small amount of evidence from which he has

constructed his arguments is not necessarily indicative of widespread feminine

participation in the Mithraic tradition. It is therefore possible to conclude that David’s

arguments are not overall as incompatible with traditional Mithraic scholarship as he

made them out to be, but it is nevertheless true that he was reacting directly to a

number of Mithraic scholars and their traditional views, working from a rather limited

body of evidence, which does not bear out his conclusions, or at least does not discredit

the traditional notion that Mithraism was generally gender-exclusive.

Alison Griffith responded most directly to David’s article, systematically

addressing each of his arguments. In fact, Griffith acknowledged she was not the first to

respond to David’s work, saying, “The conclusion that none of this evidence is

50
unequivocally Mithraic is hardly new; the aim is to put the debate to rest.” But Griffith

also sets out to find some middle ground between the vision of Mithraism as hostile to

all things feminine and the Mithraism which welcomed the feminine as a balance to the

masculine. First, however, she sets about by organizing her counterargument to many of

David’s assertions.

Beginning with texts, Griffith reminds us that both the works of Tertullian and

Porphyry cited by David are highly contentious, and/or likely not without significant bias

against Mithraism. Griffith notes that Tertullian’s account of Mithraic rites and

participation was likely influenced by the fact that Tertullian was writing to denigrate

non-Christian religions, and that the “lioness” mentioned by Porphyry is more likely to

have been “hyena,” as the manuscripts have been corrupted, and both words end similarly

50 Alison Griffith, “Completing the Picture: Women and the Female Principle in the Mithraic Cult.” Numen
53 (2006): 48.
 25

51
in Greek. Given that the existence of a lioness grade of initiation in Mithraism depends

almost entirely upon this likely corruption, the closeness of the words is frustrating. But

Griffith quickly moves on to address the connection of the Oea tomb and the S. Prisca

Mithraeum in Rome as set forth by David, pointing out that the two sites are not nearly

so contemporary as David argued, and that an earlier scholar’s comparison of similar

figures appearing in both places was at first tentative, but later abandoned by the very

52
same scholar. Overall, Griffith concludes that the Oea site is not definitively Mithraic,

meaning that any attempt at forced association between archaeological remains there and

in the Roman Mithraeum would be detrimental to our understanding both of Mithraic art

and symbolism, and to our understanding of burials in Roman North Africa. Griffith

carries on to state that the appearance of female heads or figures in certain Mithraea is

easily attributable to later vandalism or the intentional filling in of Mithraic spaces in the

53
late fourth century CE. Given that Mithraism had largely disappeared or been stamped

out by early Christians by this time, it is troublesome to attribute the appearance of

female figures in Mithraea solely to the activities of Mithraic worshippers or dedicators.

On the topic of dedicatory inscriptions, Griffith says, “A number of dedications to

Mithras by women have been cited in support of the idea of women’s participation in the

cult, but none can be unquestionably shown to be both a dedication to Mithras and by a

54
woman.” Griffith notes that there are a few cases in which it is impossible to entirely

disprove a connection between a woman dedicator and a Mithraic inscription, but asserts

51 Ibid., 51-52.
52 Ibid., 53-54. Griffith notes that the S. Prisca depictions date to the early 3 rd century, and the Oea sarcophagi
and associated paintings date to the late 3rd or even 4th century. Additionally, Vermaseren’s lack of confidence in his
comparison of figures from the two locations was glossed over in David’s writing.
53 Ibid., 57.
54 Ibid.
 26

that even in these cases, there is significant lack in clarity of the inscriptions, as

abbreviations frequently obfuscate the meaning of what would be the most vital

55
passages. Finally, Griffith acknowledges the potential for communication between

the highest grades of Mithraic initiation with the highest grades of female initiation in

cults such as those of the Magna Mater, but again concludes that the evidence is simply

not substantial enough to found a solid argument upon, and dismisses all notions that

56
such communication would signify a high-ranking women’s grade in Mithraism.

Following her reassessment of David’s evidence, Griffith tackles the idea of the

th
feminine principle in Mithraism, as many scholars did in the 20 century. However, like

those scholars, she does so without much consideration of how the feminine principle

connected to real women, choosing instead to view the feminine as the abstract binary to

the masculine principle apparent in Mithraic imagery. As much of her discussion of

imagery is heavily iconographic in nature, and not entirely distinct from earlier scholars’

considerations of the same imagery, it does not need to be discussed at length.

Interestingly, however, Griffith does argue against the notion of Mithraism being a

religion designed to suppress the feminine completely, as promoted by Gordon and

others, concluding that the religion simply dealt more in gender ambiguities than in

57
absolutes. Thus, Griffith’s vision of Mithraism is that of a religion in which the

genders are inextricably intertwined, rather than in direct competition with one another.

While it may not be possible to assign art-historical methods to all of the Mithraic

scholars in this paper, given that their work is situated more often in the disciplines of

55 Ibid., 57-62.
56 Ibid., 62-65.
57 Ibid., 75-77.
 27

archaeology, history, classics, and religion, the disciplines do at least share enough

vocabulary that I can here make a brief attempt at defining their methodologies. For the

most part, Mithraic scholars have been limited to semiotic approaches to the religion, as

the vast majority of evidence is in the form of painting and sculpture found in Mithraea.

Iconographical studies have borne the study of Mithraism aloft, and have inspired much

th
scholarly debate from the 19 century to the present, but focus on iconography has also

left a great many questions unanswered. Scholars such as Cumont and Campbell

especially employed formalist methodologies in their attempt to reconcile Roman

Mithraic imagery and tradition with what they argued were predecessor traditions in

Persian territories. For the most part, gender in Roman Mithraism has been viewed from a

structuralist perspective, with the scholarly consensus being that the feminine was not

only present, but essential to Mithraism and the Mithraic worldview. Despite a general

lack of overt concern over the question of women’s involvement in the cult’s activities

th th
throughout the 19 and early 20 century, scholars have recently revisited the question.

Working more as social historians, Gordon, David, and Griffith again tried the case of

women’s participation in Roman Mithraism, but the conclusion was that, as originally

argued by earlier scholars, women most likely did not participate in the cult. Overall

these seem to be the approaches most prevalent in the history of Mithraic scholarship.

While the study of Roman Mithraism has been interdisciplinary from its very

beginnings, limitations stemming from the nature of available evidence have long frustrated

scholarly attempts at making definitive, unimpeachable statements about the nature of the

th
shadowy cult. From its origins in the late 19 century, modern Mithraic scholarship has

been rife with disagreement on even the most fundamental aspects of the
 28

cult, including questions of its origins, its membership, and its overall position within the

broader Roman religious structure. As a result of the evidence for the cult being largely

epigraphic and archaeological, scholarly opinions have undergone constant re-evaluation

as more Mithraea and Mithraic art have appeared in excavations across Europe and the

Mediterranean. Nevertheless, certain ideas espoused by the earliest scholars of Mithraism

have persisted, not always as a result of the preponderance of evidence which would

support their conclusions, but more often due to a dearth of evidence required to refute

their ideas. In reality, much is yet to be discovered about Mithraism, and much has yet to

be proven more concretely. Certainly, questions about the origin of distinctly Roman or

Western Mithraic practices and traditions abound. If the cult is descended from Greek

cosmological traditions, then where do the Iranian names associated with the cult

originate? If Mithraism is indeed strongly Iranian even in the most Western parts of the

Roman Empire, would Mithraic adherents have recognized the dualistic iconography in

their sanctuaries symbolizing the struggles between good and evil, or would their

understanding have been lessened by physical separation from the cult’s homeland? The

most measured suggestion which can be offered as far as the cult’s origins is that they

were myriad. As such it is important to acknowledge the differences in Mithraic sites and

iconography across the Roman world not as obstructive to our goal of understanding the

truest origins of the cult, but vital pieces of evidence for the richness and versatility of the

Roman religious syncretistic machine.

 29

Chapter Two

Attributes of a Typical Mithraeum

In order to understand the extent to which regional variation is reflected in the

tangible remnants of Mithraic worship in the archaeological record, it is first necessary

to establish a checklist of what might be expected in a Mithraic space. What were the

physical structures themselves like? Where were they located relative to surrounding

structures or landscape features? What imagery and iconography appeared in the sacred

spaces? What figures appeared most often or most prominently in the cult’s

iconographical record? All of these questions are important not only in allowing modern

scholars to identify Mithraic spaces with a greater degree of certainty (especially

important as there have been some fierce debates over the identification of some sites as

Mithraic), but also in focusing in more closely on deviations from the norm, or

differences between Mithraic sites. While it would certainly be impossible to establish an

absolutely perfect set of criteria to which all Mithraea would adhere, due simply to the

sheer number which existed in the ancient Mediterranean, it should at least be possible to

identify those factors which point most strongly towards Mithraic activity. Additionally,

it would be folly to claim that the appearance of auxiliary figures atypical in the cult’s

iconography at any site should be grounds for excluding it from a catalogue of Mithraic

sites. Not only would this reflect a fundamental misunderstanding of Roman religion’s

versatility and adaptability, but it would also wrongly assume that all Mithraea across the

Roman world were in fact frequented only by Italian Romans. The depiction of local

deities is, if anything, evidence of the extent to which a religion has been successfully

syncretized into the Roman religious program.

 30

Indeed, the differences between sites across the Roman world are the very key to

unlocking many of the cult’s secrets, as written ancient sources hardly suffice even in

providing a complete picture of the cult and its activities in any one corner of the empire,

and therefore could hardly be interrogated for information on the cult’s regional

variations. As such, in this section of my thesis I will refer to Leroy Campbell’s

58
Mithraic Iconography and Ideology extensively, as I believe his work to be one of the

most accessible and concise efforts to catalogue and establish typologies for various

aspects of Mithraism and Mithraic structures, iconography, and resulting practical

differences across the Roman world. In addition, Campbell’s volume represents an

important moment in the history of Mithraic scholarship, as he was a proponent of

Cumont’s ideas of links between Western Mithraism and earlier Iranian religious

traditions. His volume was published only a few years before the first conference at

which many of Cumont’s ideas were first rejected wholesale, meaning that most of the

evidence which was used to discredit Cumont’s ideas already existed at the time

Campbell’s work was published. Therefore, this can be read as a work not written in any

haste to rebut any new ideas, but a carefully compiled and well-organized investigation

into the shadowy world of Mithraic beliefs, via a great many different sites and cult

scenes.

Campbell creates a number of typologies and distinctions which are invaluable in

answering many of the questions I raise in the introduction of this chapter, though he

focuses primarily on the iconography of Mithraic sites, rather than on questions of the

orientation or location of Mithraea. As I am writing this thesis from an art-historical

perspective, it makes a great deal of sense to follow Campbell’s iconographical study in

58 Leroy Campbell, Mithraic Iconography and Ideology. Brill, 1968.

 31

determining differences in cult practice and setting between regions. The study of the

iconography in various tauroctony scenes from Mithraea across the Roman world is not

only one of the easiest starting points in investigating the cult, but also necessary, as this

scene is the most consistent link between Mithraic sites, and by far the best possible

determining factor when the identity of a site is in question. However Campbell does not

focus only on variations of the tauroctony and figures within that scene, but also on other

scenes and figures typically (but importantly, not always) found in Mithraic spaces. The

author also helpfully attempts to divide certain variations in iconographical programs by

the regions in which they appear, and even by their chronological period, pointing out

which figures seem to have been later additions, or additions found only in certain areas,

perhaps in deference to local traditions. In the following, I will summarize the evidence

in the hope that I might begin to construct a clearer picture of a “typical” Mithraeum in a

number of regions, in anticipation of my description of specific representative sites in a

later section.

First, Campbell summarizes one of his own previous publications in which he created

a typology of tauroctony types, mostly divided by the media or physical ways in which the

scene was created, rather than by the iconography of the scene itself. These types may be

summarized as follows (see Figs. 3-8): Type I is mostly constrained to a single rectangular

field containing all elements of the scene and does not seem to be regionally bound. In type

II, a stele with a single field is used, most often in Thrace, and later examples include a frieze

on the base. Type III, most often found in the Danube, is a stack of three horizontal fields,

with the tauroctony set in the largest middle section. Type IV is noted to be a synthesis of

types II and III, found in Dacia, with the registers on a

 32

stele. Type V from Dalmatia and Pannonia has a circular field containing the tauroctony

scene, although Campbell notes that this type later merged with many of the other types.

Type VI utilizes a rectangular field like type I, but includes the depiction of an

architectural cave, arch, or vaulted construction, and is noted to be most common in the

provinces of Middle Europe. Type VII is mostly a development on VI, including

symbols on the border of the tauroctony which are not necessarily related to the scene

itself. Finally, type VIII is yet another development on type VII, with smaller series of

images on either side of the tauroctony scene, originating in Northern Italy and spreading

59
to the Rhineland. Thus it is already clear that certain habits or modes of depiction in

Mithraic spaces were regionally bound. However, it is also true that a difference in the

medium used to create the tauroctony scene is not quite sufficient evidence to argue for

the highly variable nature of the cult across space and time.

Campbell also offers a brief summary of the subtypes he observes in the many

tauroctony scenes he discusses, all of which are based around the actual depictions of the

bull and its interaction with the god Mithras in the moment of the struggle and killing. He

identifies five subtypes, many of which are again predictable by region. In the somewhat

rare subtype A, found primarily in Southeast Europe in conjunction with subtype B,

Mithras is shown kneeling on the back of a relatively small bull, which is depicted with

its legs tucked under its body (Fig 9). Subtype B is noted to be a Greco-Roman

development of A, in which Mithras appears more in profile astride a bull which is shown

prostrated rather than with folded legs. From the Hellenistic East we get subtype C,

which emphasizes Mithras’ line of sight towards the sun as he sits atop a bull struggling

59 Campbell, Mithraic Iconography. 1-2.

 33

to regain its footing. This subtype is noted for its emphasis on the snake and scorpion in

the scene, and is the most widely preserved of the subtypes. Subtype D is similar to A,

but depicts Mithras with one or both feet on the ground, holding on to a bull much more

in motion than the other subtypes, typically shown still on its back legs, but being forced

to kneel at its front as Mithras arrests its motion. Finally, the more typically German

subtype E (Fig. 10) depicts Mithras riding a bull in motion, and is sometimes synthesized

60
with subtype D. While these differences may at first seem minor, it should be kept in

mind that groups with differing interests may have wanted to depict their cult’s central

deity in various ways. While civilian worshippers may have been content to show

Mithras either as pious in the act of sacrificing a subdued or weakened bull, it would

hardly be surprising to think that an image of a virile and powerful god astride an equally

powerful animal in the middle of performing a feat of incredible strength would appeal

more to soldiers in a camp on the frontiers of the empire. Not only could an image such

as this serve as a model to which such soldiers might aspire, but it would also likely

serve to allay some doubts or fears about contact with non-Roman enemies, reminding

those soldiers of the natural order (in the Roman worldview), in which civilization and

the masculine triumphed over wilderness, animals, and the effeminacy of non-Romans.

In addition to the figures of the bull and Mithras himself, variations in the other

actors included in the tauroctony scene might hold clues as to differences in the cult’s

emphasis in separate regions. Campbell discusses the multiple Mithraic “helpers” which

appear in regionally specific tauroctony types, and delves into the different things they

symbolized. In doing so, he focuses also on the myriad ways in which all these figures

60 Ibid., 2-3.
 34

interact, not only with one another, but also with the figures of the bull and the god

himself. The author discusses four helpers in particular: the dog, the snake, the raven, and

the scorpion. It is important to consider the role of these figures in Mithraea across the

Roman world, and to understand their time and place of origin, as many carry different

connotations and represent potentially conflicting principles, depending upon whence

they originated. Were all of these animals indeed meant to be assistants or benevolent

creatures in the tauroctony scenes in which they appeared? Did they represent conflicting

principles in conflict with one another, mirroring the struggle between Mithras and the

bull? And perhaps most importantly, were these creatures Eastern in origin, or were they

Roman additions inserted to bolster the narrative used to promote the cult’s central values

and beliefs? Campbell answers many of these questions, and it is important to bear in

mind that he does disagree with Cumont on some counts, as I will note later as each

assistant is discussed individually.

Perhaps the most immediately noticeable figure in a tauroctony other than

Mithras or the bull, the dog may have been one of the earliest additions to simpler scenes

61
of the bull-slaying, often appearing to lap at the blood being spilled from the bull.

Campbell notes a number of different things which could be represented by the presence

of the dog as well as its actions, including the appearance of dogs as hunting companions

alongside Mithras in other Mithraic scenes, connections to sacrificial and/or burial rites

of other ancient Mediterranean cultures, and even a parallel to the story of Cambyses II of

62
Persia who slew the Apis bull and left its body to be eaten by a dog. However given the

fact that Cambyses’ slaying of the Apis bull was a misdeed rather than a heroic or

61 Ibid., 12.
62 Ibid., 13.
 35

generative feat, it seems unlikely that the cult would have looked to this as a model for

the inclusion of a dog in the tauroctony. Perhaps most likely is the appearance of the dog

as a friendly companion to Mithras, as the god is accompanied by hounds in other cult

imagery as he hunts. But Campbell also notes that the appearance of a three-headed

Cerberus-like dog in a Mithraic space might serve as evidence of some connection to the

Iranian deity Ahriman, whose roles were quite similar to underworld deities of the

63
Greco-Roman traditions. Thus it should be apparent that, although the dog might have

been one of the earliest additions to the Mithraic iconographic program, it is not the most

lucrative figure from which we might attempt to extract information on the origins of the

cult. Indeed, it is only when considered alongside other symbolic animals of the bull-

slaying scenes that we might begin to assemble a more cohesive idea of its exact role and

relation to Mithras himself.

After the dog, the snake is the next most noticeable of the minor figures in a

tauroctony scene. Given the negative connotations attached to serpents in the more modern

worldview, it is not entirely surprising that the snake was at first seen as an opponent of

Mithras in early Mithraic scholarship. However Campbell quickly dispels this notion,

rejecting Cumont’s idea that the serpent appeared as an agent of Ahriman to oppose Mithras

64
and his faithful companion the dog. Instead, it is noted that in many traditions the serpent

was a representation either overtly of life and its generation, given its many associations with

fertility and rebirth, or with the constants of the known world, due to its association with

65
mother/earth goddess figures. This would make sense, as one

63 Ibid., 15.
64 Ibid., 17.
65 Ibid., 16-17. It is worth noting that, while mother goddess and fertility figures are sometimes associated with
later Mithraic sites such as Dieburg and Carrawburgh, there has been no serious scholarly push to
 36

of the appeals of Mithraism as a cult may have been promises of extended life or rebirth,

symbolized neatly by the snake, capable of shedding its skin and appearing to renew its

life at will. Additionally, this would fit neatly with the story told by the snake’s position

in the tauroctony, as the author notes that it is often seen lapping either at the semen

66
flowing out of the dying bull, or like the dog, tasting the blood of the bull. In either

case, the incorporation of the bull’s vital fluids into another living thing would point to

an interest in rebirth or regeneration, possibly demonstrating some compatibility with

67
traditional Iranian beliefs concerning the cycles of life, death, and rebirth. Finally,

recalling the extent to which early Christianity and Mithraism competed with one

another, it would hardly be surprising if the representation of snakes and serpents as vile

or deceptive in the Christian tradition were in fact an attempt to discredit pagan traditions

in which snakes featured heavily, including Mithraism.

While the raven is less engaged with the bull and other animals in its appearances in

tauroctony scenes, it was perhaps used as a symbol of Mithras’ virility, related to bird

incarnations of other Indo-Iranian deities, or perhaps inserted as a symbol of Mithras’

68
connection to the sky and his potency as a dutifully generative sun deity. In either case, the

raven does not engage as actively with the other actors in the scenes, and therefore does not

demand as much consideration in Campbell’s work as the other animals. It should also be

noted that there is likely not as much opportunity for confusion in the

support Campbell’s speculations that the serpent in earlier Mithraic imagery indicated links to fertility
deities.
66 Ibid., 15.
67 Ibid., 18-19.
68 Ibid., 24-25. This also seems to be speculation on Campbell’s part, but is worth bearing in mind later
when considering the difference between military and civilian Mithraea.
 37

symbolism of a bird as there is in that of a snake, as birds have not accrued nearly

the negative stigma as that of the snake in later traditional worldviews.

Similar to the serpent, the scorpion is another animal which was originally

considered by Cumont to be hostile to Mithras and other benevolent actors in the

tauroctony. However Campbell states outright that this notion is flawed, and that the

scorpion was hardly, as Cumont suggested, poised to thwart the generative powers of the

69
bull’s testicles and semen, but was rather itself associated with production and fertility.

Rather than the Iranian symbol of antagonism that Cumont might have suggested the

scorpion represented, it was more likely a benevolent agent with another origin. In fact

Campbell continues to state directly that, “It is quite clear that the scorpion motive in

Mithraic iconography, as well as in that of Mercury, was derived from a Semitic or

70
Anatolian tradition rather than from an Iranian,” therefore denying any likelihood that

the scorpion could have been a particularly Persian symbol in origin. Given the fact that

the scorpion appears more frequently in Italian and Middle European tauroctony scenes,

71
and not at all in sites as far East as Dura-Europos, this more Western, non-Iranian

origin makes a great deal of sense. In fact, this would be another testament to the efficacy

of the Roman syncretistic tradition, drawing other more traditional symbols from the

religion at its source in the East, and picking up symbols to add to the cult as it is carried

further West to arrive in Italy.

69 Ibid., 26.
70 Ibid., 27. While some of Campbell’s statements about the symbolism of these animal assistants in
tauroctony scenes was speculative, later scholars such as, Speidel (1980): 5-6, and Ulansey (1989): 15, identify the
animals simply as products of the zodiac and cosmological arrangements that informed Mithraic art and
iconography.
71 Ibid., 25. Speidel (1980): 5 also references a monument in Sidon in which a scorpion wraps around the
scene and also interacts with those within the scene.
 38

The final element of the tauroctony which varied significantly across numerous

Mithraea was the inclusion or absence of a cave in the scene. While the origin of many

elements of the bull-slaying scene may be difficult to locate definitively, Campbell casts no

such aspersions on the motif of the cave, saying “Moreover the cave imagery was essential

to the performance of the mystery rites which, though falsely attributed to Zoroaster, had

72
their origin in Iranian religion.” Despite this, however, Campbell also notes that depictions

of the cave in tauroctony scenes of Roman Mithraea appear to have originated in Rome and

73
spread only later to more Eastern sites in Syria. The author describes the cave types as

follows: The naturalistic type (Fig. 11), appearing mostly in Italy, Sicily, and Middle

Europe, likely spread from a source in Rome. The artificial cave type, depicted as a shallow

cut made into a cliff face, is most typical of Middle Europe, North Africa, and Southeast

Europe. Less regionally bound, the architectural cave (Fig.

74
8) appeared in Middle Europe, Germany, Syria, and even South Russia. While

Campbell allows for the possibility that the depiction of caves originated in Anatolia and spread

only later to Italy and out from there, he does not espouse this idea himself, content with the idea

that the naturalistic cave found in Italian Mithraea was likely the
75
earliest appearance of the addition to the scene. In any case, it is apparent that the

reality of the cave motif’s origin runs counter to the idea that all elements of Roman

Mithraism were direct adoptions from the East only later carried west to be synthesized

into the Western version of the cult.

72 Ibid., 7.
73 Ibid.
74 Ibid., 7-8.
75 Ibid., 8.
 39

Finally, although he does not focus on the orientation of the Mithraic structures

themselves, Campbell does detail the variable orientations of elements such as the

tauroctony scenes within the structures, and the cultural and practical implications of the

directions most common in distinct regions. This is an important distinction, as the author

notes that there are at least a few instances in which the orientation of the main cult scene

did not match up completely with the orientation of the Mithraeum itself, especially in

the case of some rock cave or cliff structures. He asserts, addressing the example of a

cave Mithraeum, that, “This instance warns us that the actual orientation of a Mithraeum,

76
governed by physical necessities, might differ from the symbolic orientation within.” It

is noted that in the case of natural caves or similar rock structures, the most a founder of

a new Mithraeum could do was either accept the space available and modify it, or reject

77
it altogether. Even in cities, where cult structures would have been constructed

architecturally rather than carved out of the living rock, available space, privacy, and

proximity to any number of other structures would have limited the cult’s options.

Campbell also observes that deference to other religious guidelines within cities could

have similarly restricted the erection of a Mithraeum in the direction and layout

78
desired. However the cult demonstrated some versatility, often arranging elements such

as the zodiac, grades of initiation, and the main scene of the bull-slaying in whatever way

desired within the space allotted.

Campbell carefully types and categorizes the orientations of different Mithraea,

dividing the sites into four categories based on the orientation of the main axis of activity

76 Ibid., 50.
77 Ibid.
78 Ibid.
 40

within the shrines, sidestepping the confusion of the structures’ external orientations. His

general conclusion is that the Mithraea he examined fell into 4 different groups: those

facing approximately East (Group 1), those oriented between South East and South West

(Group 2), a third group (3) arranged between North East and North West, and finally a

79
group (4) in which the axes aligned approximately West. At last, these divisions begin

to impose some structure on the myriad different divisions and classifications Campbell

has made by iconography, type of relief, etc. in earlier sections of his work. He

summarizes the findings of the compiled information with

An examination of the available data on each of these Mithraea leads to the

following generalizations: a) Those oriented toward the east have their origin at a
comparatively early date; that is, in the second century, and they are also under
fairly strong Greek influence from the Hellenized East. b) Those oriented toward
the south, if not later in origin, are under a stronger Latin or north European
influence. c) Those oriented toward the north or northeast are under stronger
Iranian or Semitic influence. d) The comparatively few definitely oriented toward
the west are strongly Graeco-Iranian and tend to emphasize the Mithraic grades in
80
their pictorial symbolism.
In addition to this, Campbell generalizes the locations of the differently oriented Mithraea,

observing that Group 1 structures are most typically in Italy and Middle Europe, those of

Group 2 are either found in Rome, Ostia, or western Europe, and sites of Groups 3 and 4 are

81
largely similar in location, distributed across Italy and Middle or Southeastern Europe.

Thus, we are finally able to begin synthesizing the vast amount of

79 Ibid., 52.
80 Ibid. Campbell is not arguing here that the orientation of a Mithraeum alone is indicative of the influences
acting on the site, but is generalizing that these are the influences which coincide most often with each orientation, based
on analyses of the types and subtypes of the tauroctony scenes within.
81 Ibid., 54.
 41

information presented into what we might expect from Mithraea in various regions of

the Roman world.

Here I will attempt to compile the information Campbell presented into a

tentative list of features of Mithraea in distinct regions of the Roman world, progressing

from East to West. For this, all of my reference will be to the same lists provided in

82
Campbell’s book, as he compiles his classifications into one table. In the easternmost

sites identified as Roman Mithraea, in Syria and near Iran, we would expect Mithraea

likely facing north or northeast, with rectangular scenes of the tauroctony, either by

themselves or accompanied by additional painted pictures on either side, depicting

Mithras either astride or leaning against a larger, more active bull, rather than a

prostrated, smaller bull. Any caves depicted in the cult relief would most likely be

architectural, and only at sites of relatively later dates. The animal assistants may be

present, most likely the dog and the snake if any, and again only at later sites. In addition

to the figures within the tauroctony, there might appear paintings or reliefs depicting

other exploits or stories of the god Mithras, although I will return to these in a later

section as I address sites more specifically.

In Italian Mithraea, we find the most variety within a region, but this is most likely

because of the sheer number of Mithraic sites actually extant in Italy versus some of the

other less-excavated regions of the Roman imperial world (Map 1). While there is likely to

be at least one example of every type, subtype, and composition of tauroctony, every

orientation, etc., there are still some patterns which emerge when looking at the evidence

presented. In Italy we would most likely see naturalistic caves represented in

82 Ibid., 51-53.
 42

scenes of the tauroctony, especially at earlier sites here than in any other region, with

some later additions of artificial and architectural types. Additionally, there is a large

degree of variability in the depiction of the animal helpers within this region, again

obfuscating any attempt to pinpoint their origins definitively. For the most part, the

Mithraea of the region Campbell identifies as Middle Europe follow the same patterns as

those of Italy itself, likely due to its proximity to Italy and the Adriatic Sea.

In northwestern European and British Mithraea, we are most likely to find

Mithraea oriented in a southern direction, an orientation otherwise found only in Ostia

or Rome. Unfortunately Campbell omits data on the tauroctony scenes of British

Mithraea, but it should be noted the cult scenes found in Germany are mainly the more

developed and complex types, and nearly all of them depict Mithras fighting a much

more active and dynamic bull, rather than simply sacrificing a subdued animal. It is my

own observation that this holds with an even larger pattern of the scene’s depiction

across the entire Roman world.

While tauroctony scenes with smaller, less active bulls are much more common in

and around Rome, Ostia, and Italy, the sites further removed from Rome itself along

frontiers in Germany, the area of the Danube, and in Asia more often have depictions of

the bull as an active and threatening figure. In other words, it appears that there is a

difference between scenes of Mithras piously sacrificing a smaller, more domesticated

bull, and scenes of Mithras hunting, subduing, and dispatching a wilder, more aggressive

creature. While there are no definite regional divisions between types of Mithraea which

preclude any type of cult scene composure absolutely, there very well may have been

differences in interest or motivation for joining the cult amongst members of different
 43

class or occupation. In settled and more stable areas of the Roman world, such as in

Rome or the other Italian Roman cities, more members of the cult would likely have been

civilians, interested more in the fraternity and camaraderie offered to initiates of the

mysteries, and thus more content with a depiction of Mithras carrying out his

orthopractic sacrificial and generative duties, spilling the bull’s vital fluids to release its

energies into the world. In contrast, to military members of the mysteries worshipping on

the more turbulent frontiers in camp or garrison Mithraea, the narrative of Mithras

slaying a wild bull would have been galvanizing, bearing in mind the numerous cultural

implications of civilization triumphing over the wild, masculine over feminine, and, in

particular, Romans over everyone else.

Of course, it is impossible to generate one single mold into which every

Mithraeum in a given region would fit. Instead, it must be conceded that all aspects of

Mithraic practice, iconography, building habits, etc. were distributed based on factors not

necessarily relating to one another. Campbell sagely acknowledged the wide variation

within Roman Mithraism early in his text, stating

The present work should demonstrate to the careful reader that Mithraism as a

world religion was not completely uniform in its selection or use of art forms,

nor was it more uniform in its cosmology, theology or rites. On the contrary,

there were different viewpoints and emphases as well as different uses of

symbols in different parts of the Roman World and even in different Mithraea in

83
the same city or in the same Mithraeum at different periods.

83 Ibid., 4.
 44

This is important to keep in mind in conducting any research on Mithraism, as the

variations in Mithraic practice and remnants are impossible to cut apart and organize

neatly, but bleed into one another uncontrollably. However this chapter hopefully

demonstrates that it is nevertheless possible to track differences in some aspects of the

cult across time and space, in order that we might interrogate individual sites within the

regions described, and make informed observations about their adherence to, or deviation

from, regional norms. As I have asserted before, it would be a great discredit to the

power of Roman syncretism if we expressed disbelief in an adopted cult’s ability to, in

turn, adopt symbols and habits from other local traditions following its own incorporation

into the Roman religious machine.

 45

Chapter Three

Survey of Specific Mithraea

Having established that Mithraism and its associated spaces were variable

between regions of the Roman world, it is invaluable to look at representative examples

of Mithraic sites from different areas to see if they fit with the attributes we might expect

from a particular region. This is important not only in evaluating the differences

introduced to cult imagery through local traditions, but also in tracking the spread and

proliferation of certain icons in the cult’s program across regions. In other words, to what

extent was the imagery of the cult determined by local tradition and custom via local

adherents of the cult, versus broader trends in the cult transferred between sites by

travelling members of the military. Given that the number of worshippers actually able to

participate in cult activities at any given Mithraeum was limited simply by the size of the

shrine, it seems not only possible, but likely, that there may have been Mithraic shrines

of very different character set up next door to one another, some being frequented by the

native peoples of the area, and others set up by soldiers hailing from far removed corners

of the empire.

In this section I will focus on a handful of sites from various regions, starting in

the east with Syria, and moving west to the Italian peninsula (focusing particularly on

Rome and Ostia), north to Germania, and finally to the most western Mithraea in

Britannia. The sites chosen will not all be contemporaneous, but this will allow me not

only to comment on differences in the cult between regions, but also on the timeline of

developments within the cult more broadly, and their spread across the empire. Where

possible I refer to reports on specific sites for objective information and site plans, but in
 46

some cases it is necessary to refer to texts in which considerable degrees of opinion and

interpretation also appear. I aim to offer first the facts of the sites, and only afterwards

comment on the significance of each to the arguments I have detailed previously. It is

important to bear in mind, however, that no site report can provide the objective truth of a

site’s original layout and functions, and therefore I can only synthesize what information and

evidence was deemed to be worth including in the original plans and reports.

Dura-Europos

Perhaps one of the most compelling examples of regional uniqueness, the

Mithraeum at Dura-Europos in Syria likely existed in three phases, first as a room in a

private dwelling in the latter half of the second century CE, then as an improved and

enlarged shrine in the beginning of the third century, and finally as a further expanded

space near the middle of the third century, before the Persian overrun of the city in 256

84
CE. While the remaining structure of the Mithraeum no longer stands in Dura-

Europos, the niche of the cult now stands reconstructed in the Gallery of Fine Arts at

Yale University. Although the three phases of use at the site might make it difficult to

determine what was added to the Mithraeum at different periods, Cumont reported that

only changes of the second and third phases would have appeared to excavators, as the

85
first phase was largely erased in expansions carried out in the third century. We may

also be certain that no modifications were made to the space later in the third century, as

the sanctuary itself was buried or filled in in 255 CE, in anticipation of a Persian attack

84 M.I. Rostovtzeff, “The Mithraeum of Dura-Europos on the Euphrates.” Bulletin of the Associates in Fine
Arts at Yale University 9 (1939): 9.
85 Franz Cumont, “The Dura Mithraeum.” Ed. and trans. E.D. Francis. Mithraic Studies. ed. John R.
Hinnels: Manchester, 1975. 199.
 47

86
on the city. Therefore this Mithraeum is an interesting intersection of phases of the cult,

likely containing elements of an earlier, more personal version of the cult from when it

existed in a private home, situated within a space modified many decades later. Cumont

and Rostovtzeff both agree on the dates 168 CE and 170 CE for the dedications of two

different tauroctony reliefs by Palmyrene commanders in the first phase of the

Mithraeum, both of which survive today, as a result of their incorporation into the later

87
phases of the Mithraeum. Cumont reports that the Mithraeum as first excavated is

largely typical of a Mithraic structure in its general layout, with a center aisle, flanking

benches, and a cult niche on the western end (Fig. 12). However it is also noted that

unlike some other Mithraea, the floor was above ground level, and the central aisle

terminated in seven steps leading up to the elevated niche containing the tauroctony

88
reliefs. Therefore it is worth noting at this time that the orientation of the entire

Mithraeum must have been east or northeast, given the reliefs’ position at the western end

of the space. Cumont also notes that the niche was likely set into a cradle vault, and that

89
the ceiling of the cella was likely concave and made of mudbricks and plaster. It is

noted that, in the absence of any springs on near the Mithraeum, a number of wide-mouth

basins and jars set in the floor of the structure at various points likely supplied the water

90
used in the shrine. Finally, Cumont mentions a few pits in which animal bones were

found, although there is some confusion as to whether these pits were part of the structure

86 Rostovtzeff (1939): 9.
87 Rostovtzeff, 8 and Cumont, 161-2. Cumont notes that the cult therefore may have arrived in Dura with
the Palmyrene archers serving in the Roman military during Lucius Verus’s campaign in 165 CE.
88 Cumont (1975): 163.
89 Ibid., 163-4.
90 Ibid., 164.
 48

itself as a Mithraeum, or whether these were a result of fill in and around the shrine

91
after its use was discontinued.

Most importantly, Cumont describes at length the images found in the shrine

itself, both in the tauroctony niche and along the walls and the borders of the niche. First,

there is the case of the double depiction of the cult’s central bull-slaying narrative. As

mentioned before, these are attributed to the dedications by Ethpeni (or Ethpani) in 168

CE and Zenobius (or Zenobios) in 170 CE, both of these being commanders of

92
Palmyrene archery units. It is not unlikely that these surviving reliefs were simply

reused and incorporated into the niche of the rebuilt Mithraeum, but it is somewhat

strange that two reliefs of the tauroctony might appear together, one on top of the other.

It is remarkable that the Zenobius relief (Fig. 8) appears to contain depictions of

figures not associated with the cult of Mithras itself. Cumont simply identifies these as

93
Zenobius himself alongside various other members of his family. This would hardly

be remarkable in Roman religious spaces more generally, as donor portraits often

appeared in shrines and sanctuaries. Indeed they also appeared frequently in Mithraea,

along with the myriad inscriptions one would expect to find in the cases of sponsorship

and dedication of new shrines. However even a cursory observation of the plethora of

tauroctony scenes extant today will reveal that the depiction of figures directly alongside

Mithras other than his torchbearer attendants, Cautes and Cautopates, and other deities

94
and zodiacal companions is exceedingly uncommon. While it would be less surprising

91 Ibid., 165.
92 See note 87.
93 Cumont (1975): 167.
94 In his Corpus Inscriptionum et Monumentorum Religionis Mithriacae, Vermaseren published the images
of well over 100 tauroctony scenes from Mithraea all across the Roman world. The Zenobius relief is the only one
in which figures other than Cautes, Cautopates, or other deities are present within the actual tauroctony.
 49

to find these portraits in smaller panels bordering the main scene of the bull-slaying, the

figures actually stand directly in front of the bull, and nearly touch it, as if they are

witnesses of the sacrifice themselves. Nearby stands the typical canine companion of the

god, leaping as it usually does towards the wound on the bull’s neck, while the raven

appears almost perched on Mithras’s flapping robe. It is difficult to see the snake or the

scorpion in the scene, and Cumont reports that neither are present, although the editor of

95
his piece notes that the snake is, or at least once was, included in the relief. Finally, it is

noted that this relief still bears traces of the brilliant colors with which it would originally

96
have been adorned. Of course, this is another detail that is nearly impossible to detect

simply from the black-and-white photographs available.

The smaller Ethpani tauroctony is a more typical Mithraic scene, showing Mithras

alone with many of the animal attendants expected (Fig. 13). In this relief, Mithras is

shown with his knee behind the bull’s shoulder, forcing it to the ground. The raven flies

behind his head, the hound is shown leaping at the bull’s neck, and an outline appears

between the bull’s leg and the dog’s body, suggesting that a snake was later chiseled

97
off. Next to the raven is a moon crescent, while on the other side a depiction of the sun

appears almost in front of the bull’s nose. As Cumont notes, this relief was also brightly

colored, and potentially decorated with four small glass or ceramic discs in the border of

98
the scene. Finally, it should be mentioned that, different from the Greek inscription

95 Ibid., 168. Conflicting reports muddle the issue here, and in many available photographs it is nearly
impossible to be certain, but Francis asserts that the snake did indeed appear in both scenes, although it was later
chiseled off of the smaller Ethpani relief, and is simply difficult to see in the Zenobios relief, due to its position.
96 Ibid., 167.
97 Ibid., 168. See note 95.
98 Ibid., 166.
 50

99
found on the Zenobios tauroctony, the inscription on the Ethpani relief is Palmyrene.

Therefore it is clear that the worshippers at this Mithraeum were culturally diverse.

In addition to the main cult reliefs in the Dura Mithraeum, it is worth mentioning

a number of paintings also found within. While Cumont records and describes a great

number of the scenes, I focus here only on a handful, and mostly on those still best

preserved as the Mithraeum stands today at Yale. The scenes in question are also those

noted to be most atypical of representations of the deity in spaces associated with the cult,

as it remains true that the differences between sites are likely to be the most lucrative

sources of information. A number of scenes dealing with the cosmogony of Mithras and

the god’s other heroic deeds appear, but many, as Cumont notes, are well attested in other

100
Mithraea. There is a Semitic name which Cumont attaches to the creation of the

paintings in this Mithraeum, and he asserts that this is evidence that a local member of

101
the Mithraeum executed the pieces. In any case, it is now necessary to turn to the two

most unique painted scenes from the Mithraeum.

First, and perhaps most exciting for Cumont, are two painted magi (Fig. 14) who

appear on the piers flanking the niche in which the two tauroctony reliefs are mounted.

For Cumont, these figures are unmistakable. He states:

If we seek to identify these two seated Magi in more detail the names which first
come to mind are those of Zoroaster, who instituted the mysteries, and Osthanes,
most famous among his disciples and characteristically associated with him in the
West. This painting, certainly prior to 256 A.C., would then represent the earliest
known portrait of Zoroaster, although this fact obviously cannot guarantee the

99 Ibid., 162.
100 Ibid., 170-82.
101 Ibid., 169-70. Francis takes issue with the assertion that this name is concrete evidence of a local’s work, and
instead espouses the idea that this was simply a native Syrian artist rather than a native of Dura, and that he may have traveled
within the military to other garrisons along the frontiers of the Roman Empire, lending some possible explanation to the
similarities between the later Dura paintings and those of Mithraea along the Rhine and Danube frontiers, especially in Germany.
 51

validity of the likeness. The reformer of Mazdeism is here shown holding a book,
for he is the mythical legislator of Iran and the supposed author of its sacred
102
literature.
Given Cumont’s zeal in connecting all other aspects of Mithraism and its iconography

with Indo-Iranian traditions, it is hardly surprising that we are presented with this

identification of two distinctly Eastern figures. However Cumont also acknowledges in

his notes that these figures are dressed not in the garb of magi or priests, but in the robes

103
of Palmyrene aristocrats. Therefore it is somewhat more difficult to be comfortable

with this tenuous connection to Zoroastrian tradition, especially in light of the fact that

there are other Palmyrene influences in the space, meaning it is much more sensible to

acknowledge this as simply another product of the identity of the Mithraeum’s sponsors.

Cumont even states that these look like portraits of real individuals, with considerable

104
emphasis on lines of the cheek and throat. Again, this would lend more weight to the

idea that these were indeed portraits of benefactors of the Mithraeum or perhaps local

officials more than it would cement any connection to Zoroaster and Osthanes.

In addition to the two figures flanking the cult niche, Cumont reports with great

excitement the scenes of Mithras hunting various animals, some of which still survive

with the transplanted cult niche at Yale. The deity is depicted astride a horse in pursuit

of a number of animals, including several stags of some variety, a lion, and what is

identified as a boar in the bottom right corner (Fig. 15). Underneath his galloping horse

105
appears a snake, likely acting in its usual role as one of the god’s assistants. While at

102 Ibid., 184.

103 Ibid., 183. It is also noted that Mithraic priests in other places, including the Italian peninsula, would likely have
worn Eastern costumes.
104 Ibid.
105 Ibid., 187.
 52

first glance it seems as though all of the animals in front of the horse are being driven

forward and hunted, not all of them have yet been struck by arrows, leaving some

possibility that the lion in this scene is in fact one of Mithras’s companions, acting as a

106
hunting dog. This is contrasted with a scene on the opposite wall, in which the lion

takes the place of the snake running below the god’s horse, while a wild lion and several

107
gazelles flee before him, having already been struck by his arrows. Cumont offers a

number of explanations for the reason we find Mithras portrayed thusly, drawing

parallels between these frescoes and a number of scenes from Germania in which Mithras

108
appears either on horseback or as an archer. Unsurprisingly, we are also reminded that

the Avestan figure Mithra, conflated with other deities in the Avestan tradition, is

frequently associated with various feats of archery, and is often depicted with a bow in

109
hand. There is also the possibility that this depiction was an attempt to situate Mithras

110
within the popular Iranian tradition of royal hunting scenes. But perhaps the simplest

explanation is that, much like the frontier soldiers to whom scenes of Mithras dominating

and slaying a violent bull appealed, the Palmyrene archers garrisoned in Dura-Europos

111
would have enjoyed a depiction of their cult deity excelling in the same military skills.

Whatever the case may be, I will return to these ideas later when I discuss Mithraea of the

Rhine and Danube regions.

106 Ibid.
107 Ibid.
108 Ibid., 188. It should be noted, however, that Cumont does not make the connection, suggested by Francis in an
earlier note, that these similar portrayals were a result of veterans’ movements along the frontiers of the Empire, carrying with
them different conceptions of how the god might be portrayed.
109 Ibid., 189-89.
110 Ibid., 192.
111 Ibid.
 53

Ostia and Rome

Unlike at Dura-Europos, where there was only one Mithraeum to which a great

deal of attention was devoted, in Ostia and Rome, there stood a great number of

Mithraea, many of which likely remain undiscovered, or which were dismantled and

incorporated into other structures. As such I will focus not on a single Mithraeum as at

Dura-Europos, which has received much scholarly attention as a single site, but on a

number of different shrines. This will also serve to demonstrate the variety of the

Mithraea themselves, and help to dispel the notion that all Mithraic sites in a given region

would be identical. Yet again, it is not by focusing on the similarities that we might learn

more about the cult, but by emphasizing the differences. I will refer not only to a

publication on Mithraism more generally, but also to Vermaseren’s excellent Corpus

Inscriptionum et Monumentorum Religionis Mithriacae, which will appear in notes as

CIMRM, followed by the entry number, as he records individual sites, monuments, and

inscriptions each with its own individual catalogue number.

Despite my repeated insistence that differences between Mithraea are most key

in investigating the iconography and identity of the cult, it is still true that in some

regions Mithraic sites shared a great many things in common. For instance, in Ostia it is

noted that most, if not all Mithraea discovered were set up in buildings which already

112
existed. In addition, many of the Ostian Mithraea are noted to be similar in size, with

only a few falling outside the average dimensions of 30-40x13-18 feet (9-12x4-5.5

112 Dennis Groh, “The Ostian Mithraeum.” Mithraism in Ostia. Ed. Samuel Laeuchli: Northwestern University Press,
1967. 17. In this volume only 14 Mithraea are considered and used as a representative body, spanning 150 years of development
of the cult in Ostia, from the earliest Mithraeum in 160 CE to the latest, ca. 250-300 CE.
 54

113
meters). While this is likely simply due to the fact that the Mithraea were being

constructed within buildings already subject to the city’s building plan, spaces of these

dimensions would have allowed for the typical rectangular shrine layout, with a central

long axis progressing from the rear of the Mithraeum towards the front, where the cult

image and center of cult activities would have been. While the wall paintings of Dura-

Europos would have provided supplementary narrative to worshippers there, mosaics

lining the floors of many of the Ostia Mithraea may have done the same, or may have

dictated the seating positions of differently graded initiates of the cult. Vermaseren

reports the seven grades found in the mosaics (Fig. 16) at the Mithraeum of Felicissimus

as follows:

5) A small vase between a raven (l) and caduceus (r) (Corax-Mercurius); 6) Radiate
diadem in the form of a crescent; underneath it a lamp (Nymphus-Venus);
7) Helmet; above it a lance. Military bag (Miles-Mars); 8) Lightning, sistrum and
fire spade (Leo-Jupiter); 9) Falx; crescent and underneath it a star and another falx of a different
type (Perses-Luna); 10) Crown with seven rays and with bands; torch (l) and whip (r)
114
(Heliodromus-Sol); 11) Falx, Phrygian cap, staff, patera (Pater-Saturnus).
These grades would been laid in ascending order with Corax-Mercurius being the lowest,

closest to the entrance, and Pater-Saturnus the highest, and therefore the closest to the

location of the altar and the tauroctony scene at the front of the Mithraeum. Nearly all

the Mithraea at Ostia would have had benches on both opposite sides of the long axis,

115
many running the full length of the room, although not in all cases. Thus we can begin

to imagine how Mithraic worshippers would have moved through and appreciated the

spaces and the images within.

113 Ibid., 10.

114 Vermaseren, CIMRM 299. The numbers are off as the entry also describes other numbered mosaic images unrelated
to the ascending grades.
115 Groh, “The Ostian Mithraeum.” 11-13.
 55

It is also from Ostia that we get some sense of how worshippers of Mithras

divided themselves into small groups in light of the fact that there were many of them,

and many structures of their cult scattered across the city. In the same volume on Ostian

Mithraism cited before, John Schreiber argues that, at least in Ostia, “The proliferation of

Mithraea suggests two things: first, that Mithraic communities preferred to remain

relatively small, and additional sanctuaries were added as the number of adherents grew;

and second, that each sanctuary drew its adherents from a more or less compact area in its

116
immediate vicinity.” In fact this is much the same way in which other scholars have

traditionally explained the existence of vast numbers of relatively small Mithraea in areas

of the cult’s popularity, rather than fewer and larger shrines. Schreiber also notes that a

chronology of the Mithraea in Ostia reveals an interesting pattern, as he suggests there

may be some significance to the cult’s earliest appearance in the city in the quarter also

117
containing a temenos dedicated to Attis and the Magna Mater. Especially considering

the number of scholars who argue for Western/Roman Mithraism’s emergence out of

various parts of Asia Minor, this is worth remembering.

It should be apparent, from what I have written thus far, that scholarship and

reports on Mithraea in Ostia are not nearly so concerned with the scenes of the

tauroctony themselves as with discussions of the other decorations and features of the

shrines, due perhaps in part to the lack of certainty in some cases about where in Ostia the

recovered reliefs and statues originated. And indeed it is from the mosaics of Ostia that

116 John Schreiber, “The Environment of Ostian Mithraism.” Mithraism in Ostia. Ed. Samuel Laeuchli: Northwestern
University Press, 1967. 33.
117 Ibid., 38.
 56

we get the clearest representations of the Mithraic grades of initiation and their

arrangement relative to one another.

In Rome, a number of Mithraea survived below later buildings, and many

fortunately retain their associated artwork, most importantly various depictions of the

tauroctony. Vermaseren describes many in great detail in his volume. Beneath the

Basilica di San Clemente a Mithraeum was discovered in 1867, possibly dating to the

later second century, in which was found a Mithraeum of relatively standard layout, with

benches down both sides of a long axis, leading up to a cult niche at the end of the aisle

118
(Fig. 17). While it is not clear if there was a representation of the bull-slaying in the

niche of this Mithraeum, there was an altar bearing reliefs on four sides, found in pieces

119
both within the space of the shrine and directly outside (Fig. 18). It is apparent from

looking at the tauroctony on one side of this altar that it is arranged in the same way one

would expect a wall-mounted relief of the scene to be arranged. All of the animal

companions of the god are present, with the dog and snake both paying attention to the

wound on the bull’s neck, the scorpion underneath the bull, and the raven drawing

Mithras’s attention from behind. This bull is not one of the wilder, more active types, but

prostrated, with Mithras kneeling on its back as he sacrifices it. Many other Mithraea in

Rome, though constrained by their buildings and surroundings, hew very closely to the

same type, and those tauroctony scenes which survived were largely very similar.

Germany

118 Vermaseren, CIMRM 338.

119 CIMRM 339.
 57

It is important to consider the Mithraic monuments of Germany, not only

because of their position very near the frontiers of the Empire and likely military nature,

but especially because sites like Dieburg and Osterburken informed some of Cumont’s

early arguments about the nature of Mithras as a syncretized deity, due to the

comparisons he drew between Mithras’s depictions on horseback at those sites and the

hunting scenes of Mithras at Dura-Europos. Vermaseren records the Mithraic reliefs

found at these sites in great detail, lending considerable insight into the aspects of the

deity emphasized by the worshippers in the area, and revealing the complexity of the

relief panels in comparison to those found in Mithraic sites elsewhere.

In 1926, a Mithraeum was excavated in Dieburg, and Vermaseren notes that it is

largely typical of a Mithraic layout, with two benches flanking a central aisle running along

120
the long axis of the space, and likely dates to sometime before 260 CE. Most remarkably,

instead of a central cult relief depicting Mithras in the act of slaying the bull, here there is a

double-sided relief in which the tauroctony is not included. While the god is depicted

interacting with the bull (either carrying it or walking towards it with knife in hand) in

121
smaller scenes bordering the main relief, the actual slaying of the bull is not represented.

The central image of what has been deemed the front panel (Fig. 19) of the stone is instead a

hunting scene, in which Mithras appears on horseback amidst a group of hunting hounds, in

122
pursuit of an animal that might be a long-eared hare. While the prey animals of the scene

are not so numerous as in the hunting fresco at Dura-

120 CIMRM 1246. It is important to remember that this date would make the Dieburg Mithraeum very close to
contemporaneous with the shrine at Dura-Europos.
121 CIMRM 1247. This entry contains a full description of the various border scenes, both on the front and back of the
relief tablet.
122 Ibid.
 58

Europos, the god himself is still shown in much the same way, with a billowing cloak,

atop a rampant horse, and accompanied by several animal helpers. The alternative quarry

he pursues is likely just a product of region, as a scene of hunting lions would likely have

seemed quite foreign to those worshippers actually from the region of Dieburg, and

perhaps even to those soldiers who had been garrisoned there for a long period, or who

had never visited lands in which lions were more common. On the reverse side of this

relief panel a nude figure identified as Helios or Sol descends from a throne in front of a

building, while Mithras stands off to his side, possibly conflated in this representation

123
with Phaeton, one of the companions of the solar deity in many traditions (Fig. 20).

While it is possible that this Mithraeum at one point contained a central tauroctony in the

cult niche, one does not survive today, making it possible that this double-sided relief

panel was instead the focus of the cult activity in the space. Given the shrine’s relatively

late date, this might be strong evidence for the cult’s later vulnerability to syncretism as it

settled into the local traditions of those on the frontiers of the Empire.

Elsewhere in Germany, however, there were representations of Mithras actually in

the act of slaying the bull. In Osterburken, for instance, a large relief panel (Fig. 10) was

found in what was thought to be a Mithraeum, although the space was not fully excavated

124
due to the danger of flooding. The relief itself contains many of the same scenes found

on the panel from Dieburg, showing Mithras in various stages of engagement with the

125
bull, being dragged behind it, but also carrying it in other scenes. Interestingly, many

123 Ibid., Vermaseren notes here as well that Mithras was indeed conflated with Phaeton in other Mithraic imagery,
given his interactions with Helios-Sol in some stories and images, and the traditions surrounding Phaeton’s connection to the sun
god.
124 CIMRM 1291.
125 CIMRM 1292.
 59

scenes appear similar to those in Dieburg, as Mithras is shown interacting with a number

of different deities here as well, not only limited to Helios-Sol, Apollo, Jupiter, and

Saturn, but including a number of female deities as well, including Juno, Minerva,

126
Proserpina, and even Diana. While the depiction of this many deities not directly

related to a cult’s traditions is not unheard of in Roman religion, it is at least unique

within Mithraism, as Mithras is most commonly the only god, or one of very few,

depicted in the cult’s spaces. Although Vermaseren does not indicate a date for this

Mithraeum, based on the complexity of the relief and the inclusion of all of the animal

assistants in the tauroctony, along with the comparability of the auxiliary scenes with

those of the Dieburg Mithraeum, we might also speculate that this relief dates to

sometime in the third century. This chronology would also potentially explain the myriad

deities also appearing with Mithras, as with the Dieburg relief, with the more matured

cult of Mithras becoming more and more vulnerable to syncretism with local traditions

given its separation both chronologically and geographically from some of the earliest

cult sites on the Italian peninsula. Additionally, it should be noted that in the tauroctony

central to this complex panel of reliefs, a lion appears alongside the dog and snake,

although it is in a relaxed pose, not rising towards the bull’s wound as the other animals

do. The scorpion appears in its usual place near the bull’s testicles. Although the lion’s

presence does not fly in the face of the idea that soldiers moving along frontiers carried

with them new images to associate with Mithras, it does somewhat muddle the issue of

the direction of transfer of these images.

126 Ibid.
 60

Britain

Also on the borders of the Roman world, the Mithraeum at Carrawburgh along

Hadrian’s Wall in Britain provides another glimpse into the cult’s activities further from

the epicenter of Roman Italy. While not as far removed from Rome as a shrine standing

in Dura-Europos, the Carrawburgh Mithraeum represents an interesting case, as it sits

further northwest than many other sites associated with the cult, separating it not only

from Rome itself, but also from its supposed places of origin in Asia Minor or Iran. This

being the case, we might expect to find in such a remote location some of the most

dramatic departures from typical cult activity and iconography of all known Mithraea.

However it is also worth considering that the method of the cult’s transmission to Britain

likely had a great impact upon the aspects of the cult most emphasized in Carrawburgh.

In any case, a report published in 1951 on the excavations of the Mithraeum at

Carrawburgh offers great insight into the various phases of the mystery cult’s activities

127
at the military camp.

Richmond and Gillam begin by noting that the freestanding Mithraeum, built

along the natural contour of a hill rather than set into its side, does not seem to have been

128
oriented in any particular direction. While not the most remarkable aspect of the

shrine’s construction, it is worth bearing in mind that many Mithraea do seem to have

been oriented intentionally in one direction or another (at least internally), although there

are other cases in which circumstances of the surrounding area were the only factors

dictating the orientation of the cult’s structures. Additionally, the fact that the structure

does not seem to have been set deeply into the hillside suggests that the maintenance of a

127 I. A. Richmond and J. P. Gillam, The Temple of Mithras at Carrawburgh. Newcastle, 1951.
128 Ibid., 1-2.
 61

cave-like atmosphere was not of the utmost importance to those worshipping here. In any

case, the authors describe three main phases of the Mithraeum. The earliest phase of the

structure was noted to be among the smallest Mithraea ever excavated, comparable in size to

the first phase of the Dura Mithraeum, likely with room for only a dozen worshippers to

occupy the space at once, indicating that the cult likely had only limited membership when it

129
was first established. Dating of this first phase is unclear, though the authors note that the

first of three sub-phases of the second Mithraeum at the site likely dated to around 222 CE,

meaning the first Mithraeum might have been constructed in the later part of the second

130
century. The temple’s second phase is noted to have doubled the size of the shrine, with

the extension of the central aisle, and the addition of an apse on the end of the structure

opposite the door, which is assumed to have hosted the cult’s central tauroctony scene,

131
although this does not survive. A pair of statues, likely depicting Mithras’s torchbearers

Cautes and Cautopates, was located at the start of the benches flanking the central aisle of the

shrine, and likely remained in that location throughout the other sub-phases of Mithraeum II,

which were likely simpler internal modifications of the space rather than complete

132
reconstructions of the building. The authors estimate the destruction of Mithraeum II

around 297 CE, and cite fire as the likely method of the space’s destruction, although they do

133
not speculate at motive or culprit of the conflagration. The final phase of the shrine,

Mithraeum III, was built at a higher level, directly on top of the older Mithraea, and reused

some elements of the older

129 Ibid., 8-9.

130 Ibid., 28. The authors also note earlier (12) that a coin depicting Antoninus Pius was found in the context of the
Mithraeum’s second overall phase, although this does not provide any more confident date for the first phase.

131 Ibid., 9-10.

132 Ibid., 14.
133 Ibid., 27-28.
 62

space, including the statues of Cautes and Cautopates, although these are noted to have

134
been moved to new positions following some repairs. Additionally, a seated Mother-

goddess figurine is reported to have been found in the anteroom of this building, though it

is appropriately noted that it is unclear whether this was the first appearance of the

135
goddess in the Mithraeum. A number of later third century coins appear in this phase

of the building, although the authors record an absence of Constantinian coins, suggesting

136
the building’s abandonment sometime in the early fourth century.

Thus it is clear that Mithraea across the Roman world are all recognizable as

spaces related to the cult. However the differences between Mithraea abound, thus

muddying the waters of scholarship on the mysteries for over a century. While the

examples chosen in this chapter can by no means be said to be representative of the

religion in its entirety, they at least exemplify some of the things unique to Mithraea

within their regions. In the present work, it would not be feasible to delve into the

myriad variations on the Mithraic space not yet mentioned, and therefore these examples

must stand on their own for examination. This being the case, the reader is yet cautioned

against the notion that so few examples might fully encapsulate even the state of the cult

within a region. These sites were chosen as much to highlight the variable nature of the

cult and its presence in the archaeological record as to make statements about the cult

more broadly. Additionally, the following section will be as much an assessment of

Roman religion and its immense versatility as it is an inquiry into what information we

134 Ibid., 29-32.

135 Ibid., 30. It is also noted that a similar Mother-goddess figure appeared in the Mithraeum at Dieburg.
136 Ibid., 34-35.
 63

might glean from these sites about the cult’s development stretched across space

and time.
 64

Chapter Four

Assessment of Evidence

Having summarized the arguments and debates of the past century and a half of

Mithraic scholarship, established the aspects most typical of a Mithraeum’s central

iconography, and having looked at a handful of sites from across the Roman Empire in

greater detail, it is at last time to reconcile a great deal of information. Were early

scholarly ideas concerning the cult’s direct doctrinal descent from Iranian traditions

correct, or did a lack of evidence lead earlier scholars such as Cumont and Rostovtzeff to

posit flawed or presumptuous theories about a religion shrouded even in antiquity in

uncertainty and speculation? What criticisms of Cumontian Mithraism hold up under

scrutiny, and which are due the same amount of skepticism they espoused? Given the

picture constructed of an average Mithraic space in my chapter on Campbell’s work,

what might we construe from the examination of the geographically disparate sites

discussed in the previous chapter? In this chapter I set out not only to reconcile some of

the ideas of early and later Mithraic scholarship through the mediating influence of works

such as Campbell’s and Vermaseren’s, but also to draw attention to those aspects of the

cult which might be illuminated by the similarities and differences between the Mithraea

mentioned in this thesis.

First of all, it is important to test the various sites chosen for this thesis against

the list of aspects most typical of a region, as established in the chapter on Campbell’s

work. I will therefore move, as I did in synthesizing the most typical aspects of a region’s

Mithraic spaces, from east to west. While the previous section was devoted more to the

facts of the spaces as reflected in the archaeological record, I will in this chapter pause
 65

briefly on each to address the possible significance of its unique characteristics, put forth

both by those publishing originally on the sites, but also promoting my own

observations, in order to answer the following questions: Were Mithraic spaces on the

frontiers of the Empire reflective of watered-down Mithraic traditions? What does the

variability of Mithraic sites across the Roman world tell us about the “Roman-ness” of

the cult? What might we deduce about the identity of the worshippers within these spaces

based on the archaeological record? And finally, what led to the differences visible

between Mithraea closer to the Roman core and those found in the further-flung outposts

of the Empire’s periphery?

At Dura Europos we find a Mithraeum remarkable in a number of different ways.

First and foremost, it is likely that the niche in the final phase of the Mithraeum held two

reliefs of the tauroctony scene, rather than the one found in most Mithraea. While the

smaller Ethpani relief (Fig. 13) is fairly typical of the tauroctony we would expect to

find in the cult niche, bearing its Palmyrene inscription beneath a scene of Mithras

sacrificing the bull and accompanied by the snake, dog, and raven, this relief is greatly

overshadowed by the Zenobios relief (Fig. 8), both in size and in complexity. The

Zenobios relief also depicts the cult deity in the middle of the bull sacrifice alongside the

dog and snake, but includes a number of other figures within the scene of the sacrifice,

137
possibly Zenobios himself and his sons or grandsons. Given that Cautes, Cautopates,

and symbols or faces of Sol and Luna are typically the only other anthropomorphic

figures depicted as present for the killing of the bull, this is remarkable. While it would

be tempting to say that this represents a significant breach in the cult’s standards for

137 Cumont (1975): 167.

 66

iconographic depiction of the tauroctony, it may simply have been a product of one-

138
upmanship between the two Palmyrene commanders, given their proximity in date.

Alternatively, the appearance of the donor and his family alongside the god might be

reflective of a certain proud attitude persistent among the shrine’s benefactors, which

might be reconciled with the notion that the two magi (Fig. 14) painted on the piers

framing the cult niche were in fact depictions of wealthy Palmyrene benefactors of the

later phase Mithraeum, rather than depictions of the legendary Mithraic founders

Zoroaster and Osthanes. Additionally, this explanation might mesh rather well with the

idea that the painted scene in which Mithras appears as a mounted archer hunting animals

(Fig. 15) is an appeal by the local worshippers to the tradition of Iranian royal hunting

139
scenes. However Mithras’ appearance as an equestrian bowman may just as easily

have been an appeal to the archers stationed in the city, and an attempt to align

140
themselves visually with a certain aspect or remarkable feat of the deity. While it is

also possible, as Cumont suggested, that this equestrian representation was evidence of

the transmission of cult imagery from the Mithraea of Germany and Western Europe, the

lack of definite chronology of the German shrines makes it difficult to determine in

which direction the mounted hunting imagery would actually have been moving. Overall,

the Dura Mithraeum and its imagery seem to show some remarkable examples of

deference to local tradition, or to the demands of wealthy local benefactors, but otherwise

do not deviate drastically from the typical aspects of a Mithraeum.

138 Rostovtzeff, 8 and Cumont, 161-2. Both scholars agreed on the dates of 168 for the dedication by Ethpani and 170
for that of Zenobios. It is worth noting, again, that Zenobios’ relief is the only one of more than 100 tauroctony scenes published
in Vermaseren’s CIMRM that appears to include figures other than Mithras, his torchbearers, and Sol and Luna as witnesses to
the actual animal sacrifice.
139 Cumont (1975): 192.
140 Ibid.
 67

Scholarship on the Mithraea of Ostia, while not nearly so concerned with

depictions of the cult’s central scene (most likely due to the overwhelming number of

extant tauroctony scenes from across Roman Italy) still offers some insight into the cult’s

character in the city, and potentially into the cult’s identity and origins more broadly. It is

worth noting that the majority of Ostia’s Mithraea were similar in size, and all established

within existing structures, arranged as much as they could be along the typical central

axis terminating in a cult niche, flanked by benches upon which the initiates would have

141
sat. Most significant in Ostian Mithraea are the floor mosaics of the mysteries’ grades

of initiation, such as those found in the Mithraeum of Felicissimus (Fig. 16), as well as a

rather tidy pattern of expansion across the city, noted by Schreiber as having a number of

different implications. First, the distribution of shrines across Ostia likely reflects the

secretive and relatively exclusive nature of the mysteries, as Mithraea only seem to have

been added in places relatively distant from one another, suggesting that different

142
congregations preferred to remain segregated from one another. In addition, Schreiber

questions the significance of the proximity of the cult’s earliest shrines within the city to

143
spaces associated with Attis and the Magna Mater. Not only does the idea of a

connection between the Magna Mater and Mithras fit with Cumont and other scholars’

ideas of Mithraism’s origins in Asia Minor, but it would also somewhat explain the cult’s

later compatibility with mother-goddess figures in places like Dieburg and Carrawburgh.

In the German Mithraea of Osterburken and Dieburg we find both challenges to

traditional Mithraic iconography, and also similarities pointing to potential connections to

141 Groh, “The Ostian Mithraeum.” 17.

142 Schreiber, “The Environment of Ostian Mithraism.” 33.
143 Ibid., 38.
 68

Mithraic communities elsewhere. While it is remarkable that we do not find a relief

actually depicting the tauroctony in the Dieburg Mithraeum, it is impossible to say that

such a relief never existed in the first place. It is possible that the adherents of the cult in

this region were simply not interested in the traditional narrative of the tauroctony, either

as a result of prior familiarity with the story, or in deference to religious traditions of the

144
region, in which gods on horseback were not uncommon. It is particularly remarkable

that the Dieburg relief panel (Figs. 19 & 20) on which Mithras is depicted hunting small

game is double sided, with a crowded rear side of the panel on which appear a great many

scenes of the god’s exploits and interactions with other deities. Among these divine

figures appear a handful of female and distinctly Roman deities. While this may reflect

the cult’s permeability to other traditions further from the Roman center, it may also be a

product of its relative maturity at a later stage of development, as it truly hit its

syncretistic stride within the Roman religious machine. Again it is worth noting that the

Osterburken relief (Fig. 10) features a wilder bull with which Mithras struggles, rather

than a more subdued bull prostrated on the ground simply waiting to be sacrificed. This

could very well be an appeal to the more aggressive nature of military Mithraic cultists,

as opposed to the scenes found closer to the more stable areas of the Empire, such as in

Rome or Ostia, where the majority of worshippers would much more likely be civil

servants, freedmen, or other non-aristocratic citizens to whom scenes of orthopractic

sacrifice would appeal more. Finally, the appearance of a lion within the tauroctony

scene is somewhat strange, as this is not typically one of Mithras’s helpers, though it is

144 Cumont (1975): 188. Cumont even pushed the idea of a conflation between Mithras and the Germanic deity Wotan,
although Francis notes that this idea gained little traction, and was overshadowed by the potential transfer of equestrian imagery
from Mithraea in the East.
 69

also unclear exactly what the lion is doing in the scene. Whereas Francis suggested that

Cumont had overlooked the possibility of transfer of hunting and equestrian images from

145
Germany into Mithraea in the East, the presence of this lion along with the hunting

scene in Dieburg seem to suggest that transfer of imagery instead may have worked the

other way, with veterans from more Eastern frontiers carrying both ideas into Germany.

Especially given a lion’s appearance as a hunting companion in one of the Dura-Europos

paintings (Fig. 15), it would be unsurprising if this were the visual precedent for the

lion’s presence as an ally to the god elsewhere. While lion imagery is not uncommon in

Mithraea outside these regions, and is also found in cult spaces in Ostia and Rome, it is

at least worth bearing this possible connection in mind, although much more definite

chronologies for the sites would be required to prove this direction of iconographic

transfer.

Along one of the most remote frontiers of the Roman world, the Mithraeum at

Carrawburgh lends us considerable insight into the cult’s growth in popularity over time,

and also offers an example of gradual discontinuation of cult activity, rather than a

dramatic event putting an end to cult activities, as may have been the case at Dura-

146
Europos and Dieburg. While any tauroctony used at the Carrawburgh site is absent,

statues of both Cautes and Cautopates were found in the final phase of the Mithraeum,

and excavators also found several other bases in the structure’s earlier phases, upon

147
which the same statues likely stood. While Cautes and Cautopates frequently appear

145 See note 131, as Francis raises this in his notes on the posthumous Cumont work from 1975.
146 Rostovtzeff (1939): 9. It is supposed that the Dura Mithraeum was filled in and used as a part of a fortification wall
against a Persian attack in 255 CE, while Cumont and Vermaseren share the opinion that the Dieburg Mithraeum was abandoned
after German attacks in 260 CE. (CIMRM 1246)
147 Richmond and Gillam (1951): 29-32.
 70

alongside Mithras in tauroctony scenes as his torchbearers, these statues may have

escaped later destruction or looting simply because of their distance from the main altar

and niche, or because their identities were less readily apparent to those not familiar with

the cult. The Carrawburgh Mother-goddess statue found in the anteroom of the final

phase of the Mithraeum might be evidence of cult ideas transferred from Dieburg, where

a similar deity was portrayed, given that the Dieburg shrine dates to the middle of the

third century, and the third Carrawburgh shrine dates to the very end of the century and

beginning of the fourth. However it is also possible that this is unrelated, as earth mother

deities are among the oldest and most common in Europe and the Mediterranean, and

also among the most difficult to identify due to their abundance. Again it must be

remembered that Mithras appears near to Attis and the Magna Mater very early in the

second century in Ostia, meaning that association with a fertility goddess would not have

been unique to frontier cult communities, and may have been one of the oldest aspects of

the religion.

Finally, as mentioned before, the Carrawburgh Mithraeum seems to be a case of a

site which fell gradually into disuse and decay as the popularity of the cult waned in the

fourth century, rather than a case in which the cult was forcibly stamped out either by

other more dominant traditions within the Roman religious program, or by an attack from

beyond the frontier. The complete absence of Constantinian coins in the Mithraeum

suggests that the cult was abandoned at Carrawburgh by the time he rose to

148
prominence. Given the cult’s competition with early Christianity, along with the

148 Ibid., 34-35. The authors note that there were only a few later third century coins found within the space of the
Mithraeum, and there seems to be some doubt about whether they should be associated with activity in the space, or attributed to
later deposition as the sanctuary filled with water and debris.
 71

military expansions and reforms enacted by Diocletian and Constantine, it is hardly

surprising their changes coincide strongly with the military abandonment of the cult.

With overall expansions of the military under Diocletian and subsequent alterations

made under Constantine to the makeup of the armies garrisoned on the frontiers of the

149
Empire, it is entirely possible that the relative popularity of the cult here and along

other frontiers may have been affected by an influx of new soldiers into the small

Mithraic communities. Given the largely private nature of the mysteries, they may have

been abandoned or otherwise have fallen out of fashion in the face of difficulties

150
maintaining the secrecy of the cult with the arrival of new units on the frontiers. While

it would be untenable to argue that this evidence that Mithraism was actively suppressed

under Constantine, it remains possible that any disruption to the existing Mithraic

community at Carrawburgh around the time of his rise to power coupled with the

legalization of Christianity under the Edict of Milan in 313 CE finally allowed

Christianity to overtake Mithraism in popularity, leading to the mysteries simply falling

out of fashion.

In light of the evidence from these different Mithraea across the Roman Imperial

world, it is worth revisiting some of the scholarly detailed in the historiography at the

beginning of this work. First, while it is certainly the case that Cumont’s ideas

occasionally stretched the available evidence and strayed too far into speculation, many

of his notions are at least partially vindicated by later research, as Roman Mithraism

149 E. C. Nischer, “The Army Reforms of Diocletian and Constantine and Their Modifications up to the Time of the
Notitia Dignitatum.” Journal of Roman Studies 13 (1923): 1-55.
150 This is an idea which occurred to me regrettably only very late in the writing of this thesis, but it is something to
which I wish to return when I work in the future to investigate the abandonment of Mithraism across the entire Roman world by
the end of the fourth century CE.
 72

certainly drew some influence at least from Iranian and Anatolian traditions in the names

and modes of dress of the cult’s central figures. However the connections to Zoroastrian

and Avestan traditions are not tenable when the cult is examined more broadly, as

figures such as the dog and serpent are shown not to be in opposition to one another in

most cases, but rather as allies or at least neutral to one another in many scenes of the

151
tauroctony. Perhaps in partial support of Cumont’s ideas, there is the possibility that

equestrian and lion imagery from sites like Dura-Europos was carried west by soldiers

who had served on the Eastern frontiers. However this would have been much later in

the cult’s development within the Roman tradition, rather than part of its foundation, as

he originally argued. While the Mithras, Cautes, and Cautopates all dress in distinctly

eastern fashion, it may be argued that this, along with their names, was the extent of the

ties between Avestan, Zoroastrian, and Mazdean traditions with Mithraism in its Roman

form. However, given the highly variable nature of the cult, it is still impossible to assert

that this is true for all Mithraic sites of the Roman period.

While Speidel and Ulansey both espoused fascinating possibilities as to Mithras’

origin amongst hero cults of Anatolia and Greece, neither of these theories offers a

completely satisfying answer to why the cult is so poorly attested in those areas in

comparison to Italy and the rest of the Roman world (Map 1). However it is possible that

the already established hero cults in these areas are precisely what prevented Mithras

from achieving as much popularity among locals or long term transplants to the areas. It

151 See note 70: Campbell was one of the earliest scholars to dismiss the idea of the dog and the serpent appearing as
figures antagonistic to one another, and the prevailing opinion by the time of the 1971 conference at which Cumont’s ideas were
largely called into question was that these animals were essentially working in unison with one another and with Mithras, or at
the very least more interested in consuming the bull’s vital fluids than in fighting one another. Speidel (1980): 4-6 and Ulansey
(1989): 15-
16 both promote the idea that these animals simply appear because they appear in the constellations of the
night sky.
 73

is also possible that the relative stability of these regions compared to the frontiers of the

Empire from the second century onwards would account for the much lower frequency

of Mithraic sites, as there was simply less of a military presence. Certainly Ulansey’s

idea that Mithras may have been an adaptation of Perseus is more convincing than

Speidel’s Orion theory, simply because of the relative position of the constellations in the

night sky. Given the cosmological and zodiacal fascinations of the cult evident in much

of the associated imagery, it seems entirely likely that Perseus would have been chosen

over Orion for a reason as simple as this. Despite the strength of this theory, however, it

still does not account for everything associated with Mithras in the Roman mysteries,

meaning that the hero cult was simply a step along the way in the further development of

the cult as it worked its way into the Mediterranean.

Therefore it is untenable to conceive of Roman Mithraism as either entirely

dependent upon Indo-Iranian traditions as Cumont argued, or as completely independent

from those traditions as some of his more outspoken critics like Gordon suggested.

Rather, it essential to mediate between these two schools of thought, relying more on

objective collections of data from sites where possible, and paying careful attention to the

chronology of different aspects of the cult. With these types of information taken into

account, it seems most likely that Mithraism in its Roman or Western form was

descended from the remnants of Eastern Persian traditions adopted by later Anatolian

peoples, perhaps in the late first century BCE and conflated then with hero cults of the

Mediterranean. The cult seems to have made its way into Italy via Ostia and then Rome

in the first and early second centuries CE, at which point Roman traditions and
 74

conceptions of the east (accurate or not) influenced the cult before it was spread

further, borne abroad by the military.

Additional aspects of the cult picked up on various frontiers during the later

second and third centuries then travelled with veterans along the borders of the Roman

world, leading to similarities in otherwise atypical aspects of the cult being found in

regions quite removed from one another, both culturally and geographically. We might

also note the difference in civilian cult spaces from those associated with the frontiers and

military garrisons. Whereas in Ostia and other more settled areas with higher populations

overall and therefore more potential worshippers of Mithras, we find that more Mithraea

were constructed over time, while in places like Carrawburgh and Dura-Europos, cult

sites were enlarged and repurposed over time. This might also indicate a greater necessity

for secrecy and isolation from the uninitiated amongst civilian populations, given the

lower percentage of the population who were initiates of the cult, as opposed to military

cult communities, wherein the percentage of those initiated into the mysteries was likely

much higher.

It would also be a mistake to argue that the cult of Mithras cannot be conceived of

as a unified tradition in the Roman period. While there is substantial variation between

sites across the Empire, this is not a trait unique to Mithraism. Many deities within the

Roman religious tradition were conflated with local deities on a regional level, but were

still considered part of the central Roman religious complex. If anything, the cult’s

adaptability across various regions points to the strength of its core principles and

iconography, as a more dilute tradition would likely disappear entirely further from the
 75

core of the Empire, having been subsumed by local traditions as the processes

of syncretism eroded the uniqueness of cult imagery and values.

 76

Future Research

From what I have written here it should be clear that there is still much more work

for me to do in my investigation of Mithraism, as well as a great deal more mediation that

might be done between the polarized schools of thought concerning the mysteries’ origins

and significance. In the future I would like to return more critically to the many accounts

of the cult’s origins in the regions of Anatolia and Asia Minor, and attempt to determine

the point, geographically and chronologically, of the mysteries’ first establishment on the

Italian Peninsula. With such a determination made, it would finally be possible to address

more meaningfully questions of the direction of spread of later cult imagery and beliefs.

Did additions to or complications of scenes of the tauroctony originate in Rome and

spread outward in a spiral away from the Roman core, or was Rome simply the first

major hub through which new cult developments passed before becoming visible in the

cult more broadly? If developments arose within Italy, were they made by native Italians,

or by transplants from other areas of the Roman world?

In addition to questions about the origins of the cult, I would like to address the

cult’s later stages and ultimate disappearance at much greater length. What ultimately led

to the abandonment of Mithraic sites along the frontiers of the Empire? Had the cult

waned enough in popularity by the time many frontier positions suffered third and fourth

century attacks that the Mithraea were simply not deemed to be worth rebuilding, or was

there simply too much competition with Christianity following the Edict of Milan?

In order to answer these and many other questions, in the future I hope to add to

th
efforts like Campbell’s and Vermaseren’s, as their volumes were published in the mid-20

century and therefore do not include any Mithraic sites discovered since then. By
 77

adding any new Mithraic sites, icons, and inscriptions to an already formidable body of

evidence available to Mithraic scholars, any conclusions borne out by this new

material and the evidence already compiled would be strengthened considerably.

Additionally, I intend to visit as many Mithraic sites as is reasonable in order to take

updated photographs of the monuments and reliefs still existing today. While volumes

like Vermaseren’s contain a number of invaluable images and plans, any detail-

oriented discussion of Mithraic sites demands much higher resolution photos than are

currently available.

Finally, in the future I would like to look much more closely at all ancient textual

mentions of Mithras and his worship to determine if it is possible to distinguish definitely

between Mithras within Indo-Iranian traditions and within Greco-Roman traditions.

While modern scholarship is confident about the prominence of Mithras as a figure

within Roman religion by the time of the second century CE, it is considerably more

difficult to determine whether mentions of that name--or any variation on it--in the first

centuries (CE and BCE) and before do indeed relate to the god worshipped by the

Romans. In addition, I intend to delve much more into the epigraphic evidence for

Mithraic dedications in order to investigate the identities of those making the

dedications, in the hopes that this will lead me to some better sense of Mithraic

communities’ identities.
 78

Conclusion

The history of Mithraism and Mithraic scholarship is complex and uncertain,

beset on all sides by speculation and unsubstantiated theories both ancient and modern

as to the cult’s origins, significance, activities, and even its most fundamental principles.

Firsthand insider accounts of the cult and its rituals simply do not exist, unlike those that

exist for some other mystery cults, and therefore scholarship on the cult has long been

focused on either perpetuating decades-old theories about the cult, or on rejecting any

theories which came before. Mithraism is, quite simply, still an enigma. However by

focusing on the facts of the cult available in the archaeological record, I believe it is not

only possible, but necessary to mediate between the extremes of Mithraic scholarship.

While the variable nature of the cult across the Roman world can stymy efforts to

understand it as a cohesive tradition and make statements which are broadly applicable to

the cult as a single entity, there are constants within the cult, and it is on these constants

that we must rely in order to demonstrate the continuity of the cult across time and space.

We must also turn to the unique aspects of the cult in different parts of the Imperial world

if we are to understand the cult more fully, as its compatibility or incompatibility with

different local traditions offer great insight into the largely mysterious practices and

beliefs its worshippers maintained. The persistence of the tauroctony scene in cult spaces

across the Empire is one of a handful of consistencies we might utilize, but there are

other Mithraic scenes which also occur all over the Roman world, and I intend to return

to many of these in later research on the cult.

Overall, it must be acknowledged that Mithraism is unique within Roman religion

because of aspects such as its gender-exclusivity, its broad distribution across the Empire,
 79

and its fierce competition with early Christianity. However I also argue that Mithraism is

one of the single best examples of how Roman religious syncretism worked. The

mysteries of Mithras emerged from an origin unclear even to ancient scholars, swept

across most of the Imperial world, and even in its later stages began to function like just

another component of Rome’s native religious complex, settling into local traditions all

over and incorporating iconography and figures from those traditions. The tauroctony

central to the cult was accepted by civilian worshippers as a demonstration of necessary

sacrifice and the fulfillment of duty, while to military adherents, it exemplified Roman

dominance over the Empire’s enemies, civilization’s inevitable conquest of barbarianism

and wilderness, and the potency of virility. While much is yet uncertain about the cult,

its images and iconography carry unmistakable meaning even today, and attest to the

mysteries’ important position within the Roman religious tradition.

 80

1. Typical tauroctony, Rome, from Vermaseren CIMRM 435, Fig. 122.

 81

th
2. Perseus and Taurus constellations on 18 century star map, from Ulansey p. 27.
 82

3. Campbell Type I tauroctony, Sidon, Syria, from Campbell, Fig. 75.

 83

4. Campbell Type II, Bulgaria, from Campbell, Fig. 2331 A.

 84

5. Campbell Type III, Apulum, Dacia, from Campbell, Fig. 1958.

 85

6. Campbell Type IV, Romula, Dacia, from Campbell, Fig. 2171.

 86

7. Campbell Type V, Salona, Dalmatia, from Campbell, Fig. 1861.

 87

8. Campbell Type VI with architectural cave, Dura-Europos Zenobios relief, from

Vermaseren CIMRM 40, Fig. 15.
 88

9. Campbell Subtype A, Glinitschtsche, Kertsch, from Vermaseren CIMRM 11&12,

Figs. 2&3.
 89

10. Campbell Subtype E, Osterburken Tauroctony, Osterburken, from Vermaseren

CIMRM 1292, Fig. 340.
 90

11. Natural Cave Type, Nersae, Italy, from Campbell, Fig. 650.
 91

12. Dura Europos Mithraeum in situ, Syria, from Vermaseren CIMRM 34, Fig. 13.
 92

13. Dura-Europos Ethpani relief, from Vermaseren CIMRM 37, Fig. 14.
 93

14. Dura-Europos Magi paintings, from Vermaseren CIMRM 44, Figs. 22a & 22b.
 94

15. Dura-Europos Mithras hunting fresco, from Vermaseren CIMRM 52, Fig. 24.
 95

16. Felicissimus Mithraeum mosaic showing grades of initiation, Ostia, Italy, from
Vermaseren CIMRM 299, Fig. 83.
 96

17. Mithraeum beneath Basilica di San Clemente, Rome, Italy, from Vermaseren
CIMRM 338, Fig. 95.
 97

18. San Clemente altar, Rome, Italy, from Vermaseren CIMRM 339, Fig. 97.
 98

19. Front of Dieburg Mithraeum relief panel, Dieburg, Germany, from Vermaseren
CIMRM 1247, Fig. 323.
 99

20. Reverse of Dieburg Mithraeum relief panel, Dieburg, Germany, from Vermaseren
CIMRM 1247, Fig. 324.
 100

Map 1. Locations of Mithraea in the Roman Empire, from Vermaseren CIMRM.

 101

Works Cited

Beck, Roger. "The Mysteries of Mithras: A New Account of Their Genesis." The Journal
of Roman Studies 88 (1998): 115-28.
Campbell, Leroy A. Mithraic Iconography and Ideology. Leiden: Brill, 1968.

Clauss, Manfred. "Mithras: Kult und Mysterien." In The Roman Cult of Mithras: The
God and His Mysteries, Translated by Richard Gordon. New York:
Routledge, 2000.
Cumont, Franz. "The Dura Mithraeum." In Mithraic Studies, Edited by John R. Hinnels,
151-214. Translated by E D. Francis. Manchester: Manchester University Press,
1975.
David, Jonathan. "The Exclusion of Women in the Mithraic Mysteries: Ancient or
Modern?" Numen 47, no. 2 (2000): 121-41.
Geden, Alfred S. Select Passages Illustrating Mithraism. New York: MacMillan, 1925.

Gordon, R L. "Franz Cumont and the doctrines of Mithraism." In Mithraic Studies, edited
by John R. Hinnels, 215-48. Manchester: Manchester University Press, 1975.
Griffith, Alison B. “Completing the Picture: Women and the Female Principle in the
Mithraic Cult.” Numen 53, no. 1 (2006): 48-77.
Groh, Dennis. "The Ostian Mithraeum." In Mithraism in Ostia, edited by Samuel
Laeuchli, 9-21. Evanston, IL: Northwestern University Press, 1967.
Nischer, E C. "The Army Reforms of Diocletian and Constantine and Their
Modifications up to the Time of the Notitia Dignitatum." The Journal of
Roman Studies 13 (1923): 1-55.
Richmond, I A., and J P. Gillam. The Temple of Mithras at Carrawburgh. Newcastle:
The Society of Antiquaries of Newcastle Upon Tyne, 1951.

Rostovtzeff, M I. "The Mithraeum of Dura-Europos on the Euphrates." Bulletin of

the Associates in Fine Arts at Yale University 9 (1939): 3-10.
Schreiber, John. "The Environment of Ostian Mithraism." In Mithraism in Ostia, edited
by Samuel Laeuchli, 22-45. Evanston, IL: Northwestern University Press, 1967.
Speidel, Michael P. Mithras-Orion: Greek Hero and Roman Army God. Leiden: Brill,
1980.
Ulansey, David. The Origins of the Mithraic Mysteries. Oxford: Oxford University
Press, 1989.
 102

Vermaseren, M J. Corpus Inscriptionum et Monumentorum Religionis Mithriacae. Vols.

1&2. The Hague: Martinus Nijhoff, 1956 & 1960.