Marquita Volken

The development of the cult of Mithras in the western Roman Empire: a socio-archaeological perspective.


1. Introduction.

Ordinarily, we may think that a cults geographical origin belongs among those elementary facts which can be taken for granted by the relevant scholarship, and casually assumed by every discussion. The origins and early development of the cult of Mithras in the Roman Empire, however, have remained a perpetual subject of dispute. As everyone knows, the modern founder of Mithraic studies, Franz Cumont, believed that the cult was, in a strong sense, Iranian, transmitted by hellenised mages whose teachings were slowly transformed through the centuries until the cult achieved its final form in the late Hellenistic period. Unfortunately, during the century since the publication of his major work, Textes et monuments figurés relatifs au mystères de Mithra (1896-1900), the archaeological proof required to confirm the role of the hellenised magi and the transmission of their cult to the West has not been found. The tradition of scholarship that built upon Cumonts work has found it difficult to respond to this discrepancy between his model and the empirical evidence. The latter indicates that mithraea appear suddenly towards the end of the first century AD[i], seemingly without antecedents, but all conforming to a similar architectural plan. The first mithraea and inscriptions appear at the same point in time, and clearly form part of the same cult even though they are found in geographically distant and culturally distinct areas. On any account of the origins of the cult, this geographic distribution is difficult to explain.

The usual solution of the issue of how the cult was introduced into the Roman world is to appeal to the army.[ii] This explanation works well when we are dealing with the areas where the army was stationed permanently, but fails to explain the existence of mithraea in areas not occupied by the army after the end of the first century, such as the three Gauls or Dalmatia.[iii] The aspects of the cult that might be thought to appeal to the military, such as loyalty, bonding through initiation rituals, and the formation of small close-knit male groups, do not seem likely to have had much appeal outside that social context. On the other hand, the fact that the cult was established in a commercial area such as Ostia suggests that they were perceived as attractive and useful by non-military personnel. Disregarding the heavy concentration of mithraea along the Rhine-Danube Limes, obviously in relation to the army, many other mithraea seem to owe their existence to their proximity to the Roman road system. One of the problems in discussing the development of the cult is that the foundation dates of mithraea do not fall into a clear chronological and geographical pattern: the chronological distribution does not display any coherent relation to geographic location. This problem is further complicated by the difficulty of dating many mithraea with any accuracy.

An additional problem is the difficulty of calculating the number of men in the cult, at any period of its development, which hampers any examination of its growth. The small size of mithraea appears to indicate an emphasis on small groups; perhaps ten to fifteen men could fit in an average sized mithraeum, which leads us to believe that the overall number of adherents was also quite small. This underestimation of the total number of cult members then leads to a conflict with the fact that the archaeological evidence is widespread. In order to make an impact in the archaeological record, a certain amount of wealth within the cult would be necessary; such a situation is unlikely if we were dealing with a small marginal group. The quantity of archaeological evidence is too large to have been created by an organisation made up of a small population, especially taking into consideration that the inscriptions associated with the cult indicate that the members were not rich social elites.

These two facts, the sudden appearance of the cult at numerous apparently unconnected locations, and the lack of reasonable figures for the population of Mithraists at any given time, make it difficult to use the purely archaeological evidence as the basis for a coherent social history of the cult. After struggling with the archaeological record, I turned the problem on its head, and began to look at sociological models to see if it might be possible to create a model of cult dynamics that would fit the archaeological evidence. It is of course true that the construction of a hypothetical model of a specific cult, especially a mystery cult with its secrecy and initiations, is merely an exercise which can only produce an ideal type which cannot claim to be a true representation of past events. Such a model is to be understood simply as a heuristic device, a tool for interpretation, prediction, and integration of existing research data. A model of the social organisation of the cult can serve to fill in some of the gaps in the archaeological record by offering a theoretical history of the cult that can be matched up with the archaeological evidence. If the model can give an insight into the structure and growth mechanisms of the cult, then various estimates of the numbers of adherents can be proposed, though keeping in mind that any figure used is only a convenient peg to hang the model upon, as a real total calculation is not feasible. The problem of the geographical distribution may also be related to the social organisation of the cult, so the model may give a framework that allows for an explanation of the difficult geographical and chronological distribution of mithraea.

A sociological model of the cult also needs to incorporate the timeline of the cults history. With over three centuries of cult activity represented in the archaeological record, it would not be prudent to assume that the cult stayed exactly the same from beginning to end. More important, we must keep in mind that archaeological evidence from those three centuries contains only a fraction of the material that existed, and may have enormous gaps due to the non-survival of certain materials. If the cult did practise a policy of secrecy (as would be normal for an initiation type cult) then the possibility of material survival is even more restricted, thus limiting the evidence that would support the proposed history created by the sociological model. The most important imponderable here is the nature and form of Mithraic worship before the creation of the characteristic building type of the permanent mithraeum. Comparative evidence strongly suggests that a new (mystery) cult cannot be expected to leave much trace in the archaeological record during its early phases. It is only relatively recently that improved investigative methods have made it possible to recover timber buildings: a high proportion of recently discovered mithraea in the northwestern provinces turn out to have been constructed in this fashion. In timber-poor areas, however, this consideration is of less importance, and we can assume that the worship of Mithras was conducted in a relatively makeshift manner that has left few, if any, traces.

Although no written account of the doctrines nor belief-structure of the cult has survived from antiquity, careful analysis of the iconographic sources and other archaeological evidence has produced a coherent interpretation of the beliefs of the Mithraic cult.[iv] This information is indispensable in creating a sociological model of the cult and speculating how it may have functioned.

2. A social-science model for initiatory cults.

The following general description of cult dynamics based on modern social science can be applied to many types of cult. The application of social science to ancient history demands an interdisciplinary approach, and has been applied, for example by Rodney Stark, to the miraculous growth of the early Christian church.[v] The dynamics of cult behaviour and development have been studied in depth by social scientists during the past 40 years. The social mechanisms by which individuals are converted to a cult, and the pre-conditions required for propagation of cults, can be understood through models of social networks and religious economies. For example, one of the important advances in social science has been to discredit the idea that people convert to a new cult because the official or dominant religion is not serving their perceived needs.[vi] Conversion is not the result of looking for a new ideology because the current one is insufficient: it is rather a matter of bringing ones religious behavior into alignment with that of ones friends and family members.[vii] It is thus individuals investment in conformity that induces them to adopt a given cult, once their interpersonal relationship network comes to consist predominantly of cult members. A social network is made up of personal contacts with other people usually because they are the members of the same household, neighbourhood, or family, but may also include professional relations.

The two most important needs a cult must satisfy if it is to be successful, are material and psychological (spiritual). If these are not met at both the individual and organisational level, the cult will fail. Stark argues that a successful cult will provide members with direct rewards to members, satisfy desires for scarce goods, and offer compensation for unattainable goals.[viii] Direct material rewards include status, financial gain, and useful social relationships. Money and privilege are relatively scarce goods for many people, especially in the quantities they desire: in these cases the cult provides some access to these kinds of scarce resources through its structure and power base, or is in a position to offer alternatives. Religious compensations are offers made relating to goods that do not exist in this world, such as life after death or salvation of the soul. A successful cult must be able to produce sufficient rewards and cohesion among its members so that the organisational side can function and the cult can continue to propagate itself. In order to grow, a cult must access new social networks: religious movements fail when they become closed or semi-closed networks and are unable to sustain growth.

There is power in numbers, so an effective cult needs not only to get members to maintain growth, but it needs the right kind of members. A small number of highly motivated and dedicated members will pay off better for the cult than a large number of marginally interested or free rider members. Here initiatory cults have an advantage, since they admit only really motivated individuals. The process of initiation shifts the individual from the status of an outsider to that of an insider, who now has access to the collective power of the group. Further initiations or selective admission to higher levels again limits access to the cults power structure. If there are several levels of initiation, the individuals commitment is strengthened each time he passes to a higher level. Each initiation or change of position in the hierarchy can be seen as a new bond to the cult. The positive compensation for the individuals commitment is increased access to the internal power structure and rewards, plus a connection or affiliation to the entire network of cult members, which results in further rewards, prestige, and power. Initiation cults create a system of positive enforcement among a group of adherents who have all passed the test and so become bonded to each other, and to the group as a whole, emotionally, socially and, in some cases, economically. An initiation cult is founded upon the obligation to maintain secrecy in relation to the revelation(s) so vouchsafed, which increases the value of the commitment. A cult can also use negative sanctions, such as ostracism, physical punishment, or life imperilment, against individuals whose commitment is faint-hearted or who try to leave the cult.

Cults that have been created ex nihilo by a charismatic individual generally last only as long as the creator and his first group of adepts are able to maintain their belief in the cults doctrine. Such devotion or commitment to a charismatic person is difficult to routinize, or to pass on to a second generation of adherents who enter the cult after the passing of the original founding members. Either the cult withers away, or it undergoes an organisational and doctrinal re-valuation or restructuration that permits growth on a different basis, allowing new incoming members to gain material and personal rewards, thus eliminating the requirement of contact with the charismatic founding individual. On the other hand, cults may also form around a ritual, or a holy site, or even out of a fragmented existing religion, so that they are not dependent upon an individual founder.[ix]

The life expectancy of a cult depends upon its ability to maintain and increase the number of adherents. To understand how a cult performs in society, one can use the social-science model of religious economy, a simple enough term referring to the notional totality of all religious activity taking place in a society, but provocatively described by Stark in commercial terms, such as market, clients, and firms.[x] A new cult is at greatest risk in a pluralistic religious market because it is difficult to compete for a market share among a population of potential customers with many options available. Most clients will choose the most convenient religions, usually the main stream firms. New start-up religious firms have not only to compete with the established firms, they must also compete among each other for the available marginal or deviant clients, that is, those who are willing to look outside standard solutions. A new or marginal religious firm is best served by catering to a select clientele in order to be successful. By adapting and continuing to provide an attractive product through careful adjustment of the cults ideology and organisation, in keeping with changes in the political, economic and social climate, the cult can ensure itself a long working life. As in all markets, the product must be desirable and functional to maintain adherence or commitment. These claims by Stark have been criticised by historians of ancient religion because they seem to misrepresent the degree and nature of competition in a flexible, polytheistic religious system such as that of pagan antiquity. But we can perhaps retain them in a general way for the new, essentially non-civic cults of the long Hellenistic period, such as those of the Mater Magna, Isis and Sarapis, IOM Dolichenus and Mithras. These cults developed new types of non-traditional organisation together with universal, non-particularistic claims relating to individual well-being in the world, even though they were by no means all mystery cults.

One last issue relating to the introduction of a new cult into a society is its cost-effectiveness in cultural terms. As Stark points out:

People are more willing to adopt a new religion to the extent that it retains cultural continuity with conventional religion(s) with which they are already familiar.[xi]

The more elements that correspond to the current belief system, the less new material the convert has to assimilate. The re-interpretation of familiar items can give authority to the cults doctrine, appearing to reveal new truths about standard beliefs.

3. The use of social networks: a comparison between Christian and Mithraic organisation.

Before applying these general cult criteria to the Mithraic mysteries, I would like to compare it with the early Christian church order to bring out the differences in organisation and use of social networks.[xii]

Archaeologically, the cult of Mithras appears suddenly in the last quarter of the first century (AD) in several locations geographically distant from one another.[xiii] What does this imply about its date of foundation? Comparison with early Christianity is instructive. We know from the written documents that have survived in the New Testament and elsewhere that an organised Christian church, or churches, had existed since the missionary journeys of Paul, now dated between AD 50/1 and his execution in AD 64. Yet recognisably Christian artefacts are virtually absent from the archaeological record before about 180 AD.[xiv] If we did not possess the written texts, a circumstance entirely due to its historical success, we would no doubt also assume that Christianity came into existence during the mid-second century AD.

In the case both of Christianity and Mithraic cult, there needed to be a sufficient quantity of at least moderately prosperous adherents before sufficient collective wealth could be accumulated to make an impact in the archaeological record. It is hardly surprising that there is no archaeological evidence for the cult of Mithras before the end of the first century: small cults that appealed to people far below the elite would not have the financial resources or the motivation to build permanent structures in stone. It is important to remember that even the earliest known mithraea, such as those at Mainz, Heddernheim-Nida III, Pons Aeni (Pfafffenhofen am Inn) and Caesarea Maritima, appear to have been at least partly stone-built structures, which implies a considerable accumulation of wealth and self-confidence among cult-members in the last couple of decades of the first century and the early second century AD. The wealth aspect holds true even if the buildings were merely rented quarters in existing ranges of buildings (as was certainly the case at Caesarea). It must also be remembered that (a) other early mithraea may have been discovered in the past, but their initial phases could not be dated (in all these early cases this has been possible solely thanks to their fine wares); and (b) the primitive methods of early excavation, indeed, with a few notable exceptions, up to and even after the Second World War, meant that analogous structures in non-durable materials, such as those recently discovered at Krefeld-Gellep, Wiesloch, Künzing and Tienen, would probably not have been recognisable as a mithraeum, and many such structures may thus have been lost, at least in timber-rich areas. Therefore it is important to keep in mind that the first appearance of the Mithraic cult in the archaeological record is probably not the starting point of the cults existence, but rather the moment when it had sufficient wealth and motivation to create durable structures.

The geographic distribution of the two cults indicates an important difference in their modus operandi. Christianity is an urban phenomenon, with places of worship as a central place serving a local congregation. As the network extended to neighbouring urban areas, facilities for accommodating the congregations were adapted and finally built when the necessary funds were available. As the size of the local Christian social network grew, progressively larger churches were constructed. The Christians spread from one urban centre to neighbouring urban centres. The map of the distribution of cities known to have had churches reveals a slow, general movement westward from the eastern part of the Roman empire. Geographically speaking, the Christian church spread organically by means of new social network connections established in contiguous areas. By contrast, the first mithraea appear within a short period in very distant parts of the Empire.[xv] Within a period of about 20 years, four of the first known mithraea were built or at least occupied, all following a very similar architectural plan adapted from the cult dining-room of the Hellenistic world, but hundreds of kilometres apart, ranging from Caesarea Maritima in Judaea to the Wetterau limes (Heddernheim III). The decisive factor here seems to be cult-organisation, which creates a consistent cult expression, rather than dispersion from a specific geographical origin, which leads to replication of analogous structures in contiguous areas. The buildings associated with the cult of Mithras do not appear to be dependent on the cults point of geographical origin.[xvi]

As mentioned before, conversion to a cult occurs through social networks. The difference between the two cults artefactual and geographical patterns suggests that the social networks of early Christianity were very different from those of the Mithraists. At the point it first appears in the archaeological record, the cult of Mithras was clearly not using a network dependent upon a fixed local congregation, but one associated with a highly mobile social group or organisation.[xvii]

4. A model of the cult of Mithras.

A sociological model of the cult may be helpful because archaeology is by nature an incomplete science. Archaeological evidence contains only a fraction of the material that existed, so recreating a picture of the past with archaeological remains is like trying to put together a puzzle with only one fourth of the pieces. This forces us tentatively to reconstruct the missing pieces with the help of written sources, historical data, and a great deal of human ingenuity. Even so, there will still be blanks, and worse, pieces that are difficult to interpret and fit into the reconstructed areas. For the Mithraic cult, we have only the durable material evidence and little help from written and historical sources. In this case, a model of how the cult and its members were organised may help us to recreate the development of the cult and understand how the geographically disparate sites in the archaeological record were established.

The archaeological evidence in the form of mithraea points to a cellular or modular organisation of groups of men. The iconographical information indicates that there was a hierarchical grade system accessed through initiations. The consistence of cult expression in geographically distant sites points to a well-developed modular organisation capable of maintaining and supporting conformity in the structures used by cult members. A cult that uses small groups of men organised into cells could use a vertical structure of grades for the cells and a horizontal structure for an all encompassing hierarchy, thus eliminating a central figure of authority and central administration and so creating a flexible and adaptable network. This would mean that the relations between members and their positions in the grade system were the most important parts of the overall organisation. The cult authority would reside in the enforcement of the grade system and the strength of the interpersonal relations and commitment between members.

A vertical module would consist of a single cell of members, representing adherents from the lowest grade, Corax, to the highest, Pater, and brought together by their immediate social network. A horizontal structure would consist of all the members of the same grade throughout the entire cult, with an implied or virtual relation to all other members of the same grade.

Vertical movement (advancing up the grade structure) within the cell is controlled by initiation rituals and perhaps the fulfilment of other ritual, financial and moral requirements. Additionally, advancement could have been negotiated through the social network by positive personal contact between members. A problem with vertical advancement is that the members of a given grade are in direct competition for openings to the next higher level (it seems unlikely that such openings were unrestricted, since that would have devalued them, and thus have reduced the motivation to attain them). Such competition can have a negative effect on group cohesion: it is actually a method used by pyramid commercial cults to eliminate members who do not show sufficiently high motivation. As the mysteries of Mithras used initiation rituals to select only highly motivated individuals, the problem of horizontal peer-group competition was probably solved by limiting the size of the modules or cells, so that competition between members of the same grade would have been minimised. The Virunum album seems to provide an illustration of this process at work.[xviii] This size limitation would probably have had a positive effect for the personal contact between members of the same grade who were not from the same cell.

Horizontal movement in the cult would mean moving from one cell to another. This implies the possibility of entering a new cell at a specific grade obtained elsewhere, as is probably the case with Trebius Alfius and C. Flavius Nectareus, each of whom seems to have joined the Virunum mithraeum with the initial grade of Pater, in the years AD 183 and 190 respectively.[xix] For example, a member of the grade miles presumably would have been able to leave his original cell and join another cell while maintaining his accredited grade. This kind of open-door policy can only work when there is a high level of confidence in the initiation systems ability to produce trustworthy adherents, and we know that honesty and fidelity are two of the virtues extolled by the Mithraists. The confidence level needed to make such transfers work is probably due to the initiation process which creates a basic measure of an individuals commitment to his cell or group. If that rite of initiation is accepted as common currency among all the members of the cult, even among members who have not previously had direct personal contact, then individual movement between cells would have been feasible. The use of an accepted value for a members commitment through out the entire cult structure would have been a necessary part of the cults structure, perhaps a feature needed for accommodating members that were part of a mobile structure like the army.

The crucial point for determining the value of a members commitment was the initiation into the cult. The scenes painted along the fronts of the podia in the Capua mithraeum around AD 220-40 give some idea of the admission ritual.[xx] Further information is provided by the scenes in barbotine technique on the recently published Mainz krater.[xxi] The early date (ca. AD 120-40) of this vessel provides some reason for thinking that the initiation process was already established and even visually codified by the time of the cults first appearance in the European archaeological context.[xxii] The most important scene in the present context depicts a young man (the initiand), in a scanty tunic, who is raising his arms, which are apparently bound together, in front of his face, as though in fear, and a second man, much more substantially dressed (the initiands mystagogue), who is raising his right hand in a gesture denoting I have something important to declare; and a third man in a Phrygian cap (evidently the Pater), seated on a chair, aiming an armed bow at the terrified initiate. All these images suggest how the initiation process worked. A person would be introduced to the cell by a sponsor, an existing member of the cult, and be accepted as a candidate for initiation. After full filling the necessary requirements (whose precise nature is unknown), the initiand would pass a first initiation, still under the protection of his sponsor. Presumably there were initiations into each grade, but it is not known whether there were further sub-divisions within the grades. To raise the value of commitment to the cult, the initiation procedures had to be difficult, even life threatening. The violence and the emotional distress of initiation-rituals in the cult have been pointed out by Beck.[xxiii] Undergoing such an experience creates strong bonds between the members.

The small size of most mithraea implies that a limited, close knit group was seen as model. However, if a cult is to be successful, it needs to be able to integrate new members and expand into new social networks while maintaining substantial continuity of structure and belief. To understand how new members entered the cult, a model of organisational dynamics can be used. The first question to be answered relates to the minimum and maximum number of men in a cell. On the basis of the conventional understanding of the sevenfold grade-structure, it would seem that seven individuals, one occupying each grade, would constitute a minimum number of men. But the arrangement of podia in mithraea does not seem to reflect an organisation based on the number seven. The podia are often perhaps usually -- of equal lengths, along the sides of the room or building, which indicates a roughly equal distribution of members on each side. But seven men cannot be divided evenly by two (unless we assume that one member, say the Corax, did not recline but served). The direct relation between the cosmic topography of a mithraeum and the podia suggests that the members would be equally divided between them.[xxiv] In the Mithraeum of the Seven Spheres at Ostia, for example, each bench is divided into six sections and each section is associated with an astrological sign. If one man was assigned to each sign, then there would be six men on each bench, giving a total of twelve. A man in reclining position needs a space roughly 1.40 1.50 by 0.80 to1.00 m, so the minimum room required for six men would require a podium size of 4.80 by 1.40 m. Two podia of 5 m. in length would therefore provide sufficient space for twelve men in reclining position, aligned one after another, feet towards the walls, heads towards the central aisle, most probably looking in the direction of the cult niche. The smallest mithraea conform to this minimum size, confirming that a minimum number of twelve men divided among two podia is probably the smallest size of a cell.

But how is one to fit the occupants of seven grades into a minimum number of twelve men? I would suggest that the reclining positions were filled in the following manner: one Pater and one Heliodromus, doubling as representations of Mithras and Sol, followed by two Persians, doubling as the torch-bearers Cautes and Cautopates, and then with each of the subsequent grades being represented by two men, so two Lions, two Miles, two Nymphs, and two Coraces. This gives a total of twelve men, each one occupying a space aligned against a different astrological sign. The grades with two men may have been differentiated through the associated astrological sign, so that the grade of Lion/Gemini would not be the same position as Lion/Capricorn.

Table 1, Grades and astrological signs, plus planets, arranged in the order presented at Sette Sfere ( after Gordon, 1996).

rising spring Sun


Pater Aries


Heliodromus Pisces


falling autumn Moon



Persian Taurus


Persian Aquarius




Lion Gemini


Lion Capricorn








Miles Cancer


Miles Sagittarius








Nymph Leo


Nymph Scorpio




Corax Virgo


Corax Libra



But this hypothetical scheme does not provide any information about the total number of men who were members of the cell, it provides a purely ideal or static picture containing only twelve men, and does not explain how new members could be integrated. The growth of the cult of Mithras is an historical fact: many, if by no means all, mithraea had to be increased in size at least once in their lifetimes as cult-buildings. We can conclude from the dimensions of the largest mithraea, such as those at Mainz (if Huld-Zetsches hypotheses are correct), the palazzo imperiale and the Mitreo degli animali at Ostia (V. 250; 278), Spoleto (V. 673), or the temple Vermaseren calls Carnuntum III (V.1682), that many more than twelve men could have participated in a given banqueting ritual.[xxv] A podium 20 m in length could accommodate 24 to 25 men, giving a total of 48 -50 men present simultaneously. So in order to allow growth, our model must to be able to accommodate more than one or two individuals for each grade. From the geographically distant placement of mithraea, the cult must have had well-established growth mechanisms that allowed each cell to expand to a certain size and then divide into new cells maintaining the grade structure in order to have created such a wide dispersal pattern.

The organisation of the grade system is fairly well understood through inscriptions and images.[xxvi] By applying a model of the cult into the known grade structure, we can model a cells size mathematically using a pyramidal structure. While there is not any concrete evidence that the Mithraic cult used a pyramidal structure, this kind of organisation is the most successful and stable and fits well with an initiation type cult. Such a model needs to be able to accommodate new members while maintaining a roughly pyramidal structure a classic exponential pyramid scheme is implausible, since it would imply the following numbers:

1 Pater

2 Heliodromus

4 Persian

8 Leo

16 Miles,

32 Nymphus

64 Corax.

 This would mean that the lowest group, Corax, would be too large to allow close personal contact both for the horizontal and vertical structure.. Moreover the reduction to a single person at the top suggested by the above exponential model does not fit well with a system of cells that appeared to have had members that could move from one cell to another. The reason is that, in the classical pyramid scheme, each member is directly linked with a superior member and has control only over those junior members directly linked to him. Another objection is that, if we assume that the numbers of places in each grade were limited as in the above classic pyramid scheme, then the integration of new members would become impossible after the cell had reached 127 members. Classical pyramid schemes deal with growth by having an unlimited number of levels, creating a vertical power strategy with a single person at the top of an unlimited number of levels. For the Mithraic cult, the seven grades create a limited number of levels in a pyramid structure, indicating that the growth mechanism has more horizontal volume than vertical. The creation of new cells through the division of existing large cells would be the logical manner to accommodate a growing number of adherents.

I would suggest that the growth mechanism relied, in part, on the close relations established between the sponsor and his initiands. For, as is suggested by the mystagogue-figure who accompanies the initiand on several of the Capuan panels, and in the scene of the supposed initiation of the Corax on the Mainz vessel, there apparently was a great deal of direct, personal control over the initiand. Such rituals act as a mechanism to control behaviour and commitment. Such intimate personal contact is a usual, probably indispensable, feature of a close and durable social network. Members may have had an affiliation to their sponsor that helped to decide which members went into the new cell when the original cell became too large. This relationship may have mimicked the client patron system well known in Roman society. Such replication of social experience is exactly what we would expect to find in a cult, which needs to be, in Starks terms, culturally cost-effective. Gordon long ago argued that the cult of Mithras mirrored Roman social experience, and saw the cult as a confirmation of ordinary social experience.[xxvii] Although, or just because, the cult of Mithras does not mirror official Roman religion, it needed some form of cultural familiarity to have been attractive to potential adherents without their being required to invest an unrealistic amount of effort in new learning.

The movement of members from one cell to another along the horizontal structure would, as mentioned above, be assured by the accepted value of the initiation, allowing a member to be integrated into the vertical structure under a new patron. This horizontal movement may mean that the relation between sponsor and initiand (in imitation of the client/patron relationship) was less strong than horizontal bonds within the grades, implying that the doctrine and authority of the cult served an important role in unifying the overall structure of the cult. Concretely, this would mean that an initiand would have had a sponsor who could have been in any level of the grade system, but the initiand would enter the cult at the level of Corax, and thus come more directly under the influence of the peer-group rather than of his sponsor. It would not make much sense for a member of the lowest level to be able to sponsor other new initiands, but it may have been possible. The question of when a member could become a sponsor is an interesting one, and would clarify a great deal how the growth mechanism worked, but is a question that cannot be answered. A layering of personal networks, administrational structures and religious doctrine would have given the cult more opportunity to provide the members with the rewards associated with prestige by presenting a multitude of organisational positions.

We do not know how quickly an adherent might move up the ladder of grades, but, given the complexity of the knowledge or insights to be assimilated, the moral character to be acquired, and the financial outlay involved, the rate of advancement is unlikely to have been rapid. The higher the grade, the more commitment was required from the cult-member, so that the amount of financial and other commitment demanded at the highest levels may also have served to slow the upward movement of cult members. It is also probable that there were some members who did not desire to move up in the structure, and were content to stay in a certain grade for many years. Again, the long-term growth of the cults development must taken into account. When the cult was new, the cells were small, even tiny, and advancement relatively easy; but as time passed the popularity of the cult presumably produced an excess of adherents, and consequently more obstacles to advancement were needed in order to increase the value of the commitment. These same obstacles were relaxed when the cult was no longer so popular, as a means of maintaining membership. We do not know how promotion from one grade to another was effected, but there certainly had to be an accepted and fair method, and the long success of the cult points to a system that worked well.

As I have argued above, the classic exponential pyramid is not a suitable model for the cult of Mithras, as we know that a system of cells using a pyramid structure of only 7 levels was used. While we do not know how many men occupied each grade, we can start with a small model using the structure of grades in table 1, and add members to show growth. New members enter the cult in the lowest level of Corax and were probably servers. The need for secrecy would imply that the servers were members of the cult and not uninitiated serving slaves. For administrative needs, there may have been a chief of each grade, and it may have been only the chief of each grade who was seated on the podium during banquets, especially in the smaller mithraea.[xxviii]

Table 2, Men occupying the grade system, columns 1 showing the places and grades aligned with the podia on each side of a mithraeum, columns 2 showing chiefs or administrative heads, columns 3 showing a pyramidal distribution of members in the grades.

























Lion (Capricorn/Venus)



























The mathematical model presented in table 2 gives a total cell size of 42 men. The division of the grades Lion, Miles, Nymph, and Corax into two subgroups allows a larger number of men to be in the grade while maintaining close personal control over the members. This would mean that, on entering the cult, a new initiand would have a direct personal relation to his sponsor, who would be in a superior grade, while also being subordinate to the administrative head of the Corax grade, and ultimately under the control of the Pater. In this way, the member is connected to the cult through personal, administrative and doctrinal bonds, in addition to his initiation.

A tabular model of growth can be estimated by allowing a steady rate of new members and upward movement in the grade levels. This model is only useful for seeing the possible structure of growth and does not take into account horizontal movement of members among cells.

Table 3. Growth table based on a cell with members arranged as in table 2, with new members entering through the grade of Corax and members moving up to higher grades (in rows 1-6). Numbers in bold indicate an administrative head while numbers in parentheses show the division of members between the head(s) of a grade.[xxix]













2 (1)(1)

2 (2)(2)

2 (3)(3)

2 (4) (4)

2 (5) (5)





2 (1)(2)

2 (3)(3)

2 (4)(4)

2 (5)(5)

2 (6)(6)




1 (1)

2 (2)(3)

2 (4)(4)

2 (5)(5)

2 (6)(6)

2 (7)(7)



1 (1)

1 (3)

2 (4)(4)

2 (5)(5)

2 (6)(6)

2 (7)(7)

2 (8)(8)



1 (2)

1 (3)

2 (4)(4)

2 (5)(5)

2 (7)(7)

2 (8)(8)

2 (9)(9)



The cell size in row 5 of table 3 gives 83 men, who obviously could not all fit at the same time into the majority of mithraea, but perhaps they did not need to. If the mithraeum was the place where the rites took place, perhaps only a proportion of members of each grade needed to be present inside. The evidence of Tienen (Belgium), where a grand feast of at least 100 people was held one summer around AD 275 outside the (small) mithraeum, apparently to celebrate the repair of the building, shows us that the number of men participating in the feast largely exceeded the capacity of the mithraeum[xxx]. By the time that the cell had reached the size shown in row 5 of table 3, the lowest group of Corax would be rather large for close personal contact, while the grade of Pater has three members, a figure which seems large for a single cell.[xxxi] When the cell had reached the size of 76 or 83 men, it would seem logical that it would be divided into two smaller yet complete cells in order to accommodate further growth while maintaining the necessary close personal contact. The two new cells could probably use the same mithraeum if there were no strict calendar of mandatory ritual banquets. As we do not know the reasons and frequency for celebrating the ritual banquet, it is not possible to speculate further about how and when the mithraeum would be used.

A mithraeum was certainly needed to house cult images, and was evidently the place where a ritual banquet took place. Complex initiatory cults generally require the members to learn a great deal about its structure and belief system. The information or knowledge presented at the higher levels is obscured from the lower levels, access to new information or deeper truths being one of the rewards offered for commitment. Lower grades are not privileged to know the most sacred of truths presented by the cult: as Apuleius suggests in the Isis-book of The Golden Ass, the revelation of the next level of knowledge can only be attained through personal, including financial, engagement. A period of preparation for the initiation into the next grade would have been required. The mithraeum itself may even have served as a classroom for preparing the initiation to a higher grade. The training and preparation of each man who wished to progress to a higher grade could have taken place in the mithraeum, using the cult-icon as a teaching support. The individual grades can be understood as forming micro-cells within the vertical structure of the entire community. As the example of the Patres patrum at Dura-Europos indicates, such a structure could have also been extended to the horizontal structure, i.e. a macro-cell consisting of members of the same grade from different cells.

The mathematical model in Table 3 is of course an idealised picture of growth, but it gives us a starting place for considering how mithraea could have been organised. It is quite possible that some cells were very stable once they reached their maximum size. This would have been a normal result when the cell was established within a stable social network and absorbed all of the potential members. The short-term success of the individual cell is assured, but not the long-term success of the overall cult. In the situation where men were not bound to a geographical location, but were constantly being moved every few years, each time they moved they would come into new social networks and then there would be new openings to be filled in a cell. Taking this mobility a step further, it is possible that cells may have divided and re-form with members from other cells that already existed in the area and from displaced members from other cells. This kind of cell formation would be analogous to the use of vexillations in the army, where men from different units were assigned to a temporary formation in order to perform specific tasks in areas where their unit was not stationed. On this analogy, a Mithraic temple might sometimes have been a temporary structure needed only for short time, given that the members may never have returned to that area after moving on. If the men who had been temporarily stationed in an area were established long enough to connect to a local sedentary social network, then a cell may have been formed in that place and would be maintained by the local non mobile members. This would have been advantageous both for the mobile members and the sedentary local members. The local members would have new contacts through the cult, and future members who would come into the area would have an already established cell of their cult available, plus a mithraeum already constructed and in use.

This model suggests how the cult could have been organised into cells that primarily made use of a horizontal strategy for growth by dividing the cells to create many small pyramidal structures. The complexity of personal, administrative, and religious relations ensured a stable organisation even when the cult members moved from one cell to another. This mobility of members either in a geographical or organisational sense gave the cult and its members practically unlimited possibilities for expansion into existing and new social networks.

5. The relation between cult numbers and mithraea

Let us now go back to the first mithraea that appear in the archaeological record. Beck has suggested that a small founding group entered the Roman world from Commagene as late as the third quarter of the first century AD.[xxxii] Such an initial group would have needed very powerful and deliberate conversion methods to have created any archaeologically-visible impact in the Roman world by the end of the century: the principal objection is that rapid conversion growth rates are statistically rare among cults. One exception is the Mormon Church, one of the fastest growing cults in the modern world. By means of an effective conversion strategy along established social networks, obligatory high birth rate among members, and few restrictions for new members, this sect has managed an historical rate of increase of 43 % per decade during the twentieth century.[xxxiii] As a restrictive initiatory cult which excluded women, and within the entirely different world of antiquity, the cult of Mithras could not possibly have had anything approaching such a high rate of sustained growth.

But, we can test Becks date by making two projections, one assuming an initial group of 100 and the other of 1,000, increasing at a rate of 40 % per decade, and starting in the year AD 70.[xxxiv] This would give the following population figures:

 Table 4. Cult numbers based on a 40% growth per decade, starting with 100, and 1,000 members.










No. of

Men in founding group

















 Could a cult consisting of 196 men have made such an impact on Roman society by around AD 90 that the poet Statius (Theb. I.719-20) could refer to the god Mithras in the expectation of being understood by his audience? By increasing the number of men in the founding group from one hundred to a thousand, with a 40% rate of increase per decade, the sum of 2,744 members in the year 90 presents a more significant group, but not when we consider that the cult was most probably spread over the Roman empire. Even with a large founding group of 1,000 members, reaching a population of 7,537 in the year 120, it is hard to say if this is an adequately large number. Could seven and a half thousand men, spread over the empire, be sufficient to make an impact on the archaeological record? It was not sufficient for the Christians; Stark estimates the Christian population at 7,530 in the year 100.[xxxv] In Rome, the earliest catacombs date from ca AD 150; the great bulk of Christian archaeology dates from after AD 250, indeed from after Constantine.

Even if we increase the base numbers in this way, the total number of adherents in AD 100 seems too small to have produced the known archaeological evidence. There is also the problem of the similarity of the grade structure with the army. What would be the cultural advantage for a founding group composed of Commagene court officials based in Rome in using a hierarchical structure analogous to that found in the Roman army? The years around AD 70 do, as Beck has pointed out, seem to be significant in the cults development, but in my view it is incredible that a cult that only began in the Flavian period could in such a short time have accumulated the numbers of adherents and the cultural adaptations implied by the archaeological evidence - especially a secret cult that used violent initiations, and that was not linked to the rich social elite.

The scenario looks more plausible however if we shift the founding group back in time, to the previous favourite candidate, the Cilician pirates of Plutarch, Vit. Pomp.24.[xxxvi] If we assume a maximum group of 1,000 cult members in the Roman world in 60 BC, and a growth rate of 23% per decade, we obtain the following figures[xxxvii]:

 Table 5. Estimated growth starting with 1,000 members in the founding group at a rate of 23% per decade:

Year BC AD Number of members


60 1,000

40 1,512

20 2,286

 01 2,811

 20 4,252

 40 6,431

 60 9,729

 80 14,718

 100 23,266

 120 27,387

140 41,433

If we assume that the cult spread first within the army, there would have been sufficient time for it to imitate a military hierarchy, develop a range of initiation rituals, create a codified iconography and a standardised architecture by the time of the earliest archaeological evidence. On the hypothesis of a growth-rate of 23% per decade, the cult would have had around fifty thousand members by the middle of the second century AD. While the rate of increase of most cults slows over time, as the potential pool of interested persons is absorbed into a cult, access to new social networks can offset a growth decline for a certain period of time. While there is not any simple way to quantify the effect on the growth rate, the fact that the cult was open only to men and that the population of men was statistically superior due to female infanticide must also speak in favour of a reasonable growth rate. But even the numbers in Table 5 are not large enough to approach Becks estimate of the cults membership in the Severan period of some 1,200,000 adherents, that is 2% of the hypothetical total Roman population of 60 million.[xxxviii] A constant growth rate of 23% with a founding group of 1,000 in BC 60 only gives 384,431 members for the beginning of the Severan period.

 Historical events may also have sporadically affected the growth rate once the cult was in an established network. For example, the crisis of AD 69 may have had an effect within the army: those who belonged to the cult of Mithras had a system of trust among each other which must have been advantageous during that time of crisis and loss of leadership. This may have created an upsurge in conversions and increased the overall number of members. The period around the year AD 70 may, as Beck has suggested, have been important for a different reason, namely as the period when the cults astrological aspects were developed through the influence of Ti. Claudius Balbillus.[xxxix]

 But even on the conservative estimate of a founding group of 1,000 members, there is still the problem of the lack of archaeological evidence before the end of the first century AD. There are too few known mithraea, and they date from the end of the century. There are two main considerations here. The first is the archaeological recognisability of mithraea built of non-durable materials. On the one hand, a mithraeum constructed of wood is not easy to identify as such. A trench cut in the earth, baulked up with wood or sods to construct the podia and covered with a temporary roof, would in an average archaeological context offer hardly more than rows of postholes with perhaps a small amount of ceramic and faunal finds.[xl] It is only recently, with aggressive light-industrial building on green-field sites away from the core areas of Roman settlements that such temples have been found, especially in Germany, in considerable numbers.

 The second consideration is the issue of temporary and/or non-specialised versus permanent and/or specialised places of worship. In its early years, as it certainly did later, the cult may have used rooms in insulae (such as the Casa di Diana at Ostia), or temporary quarters in large buildings, such as baths or horrea, or structures in non durable materials. These quarters would simply have been given up when no longer required. This would be intelligible for a cult using a mobile social network and who were committed to secrecy. The decision to build permanent (stone) mithraea may be explained as the result of the acquisition of sufficient numbers, and so a wealth platform, and/or of Romanisation, or it may reflect a reduction in the cults need for secrecy. The most plausible explanation is doubtless pragmatic: for a mobile cult using temporary structures, there would come a point at which it would be more efficient to build a stone mithraeum for long-term use - many individual adherents might be passing through the area, or be stationed nearby for short periods. That is, at some point it would become sensible to develop a building strategy for present and future use rather than constantly build temporary structures each time one was needed. A stone mithraeum could then be seen as a permanent substitute for a temporary wooden structure.[xli]

6. Conclusion

The model shows that the cult relied more upon the stability of its organisation and trustworthy personal networks than on a central authority housed in a permanent structure. The hardest thing to see in archaeological remains is relations between people: in the case of a cult based on secrecy and initiations, it is not surprising that it should not start to make an impact in the archaeological record until late in its history. From the model, we can understand that some mithraea were temporary structures: not the permanent centre of a hierarchically organised cult, but merely structures that housed a proportion of the cell members for a ritual banquet.

A successful cult creates a social structure that gives its members material and psychological benefits. As we have seen, the psychological implications of a social network made up of men bonded through initiation rituals may have been very important in the cults success. Organising men into cells with further sub-divisions into grades allowed close personal contact between, and control of, cult members. The absence of a central point meant that the cult was able to profit in its growth from its members relatively far-flung social contacts. The resulting numerical strength in turn made it possible, and plausible, for adherents to cash in their social contacts not only for their own benefit but also for that of other adherents in the same cell. In a society based on patronage and personal face-to-face contact, the conjunction of this- and otherworldly salvation was an important incentive to adherence. Especially in the army, where men were frequently transferred to other locations, one could be assured of having significant social contacts through the medium of the cult wherever one found oneself.

The aspect of material reward may have been linked to the power base that resulted from the social network created by the cult. Such a movement would need to have provided sufficient material rewards to satisfy the personal desires of the members. In my view, this could have been easily accomplished through commercial activity using the social network. I am not suggesting that the mysteries of Mithras were an economic cult, as the term is now understood, rather that it must have had material rewards that consisted of tangible financial benefits. People are attracted to a cult that promises earthly rewards when they believe that they merit them but are not receiving them in the desired quantity.[xlii] From the epigraphic evidence, we know that it was not the wealthy social elites who were attracted to the cult, but rather the middling economic class, the army, the patrimonial bureaucracy, and the local sub-elites who were not wealthy enough to move in the upper circles but had hopes of improving their lot. These groups fit the profile of people who believe they deserve more but are not getting it.

The development of a hypothetical model of the growth of the cult of Mithras may not have solved any of the problems inherent in Mithraic studies, but it does suggest the possibilities offered by a social-science approach to ancient history. The problem of the widely scattered geographical locations of Mithraic finds can be seen as the physical remains of a widely spread mobile cult that lacked a central authority. The small size of mithraea is not a conclusive indication of the total number of members in a cell, since the total population is rather a function of the organisational structure. While my estimates of the numbers of Mithraists in the early years are probably too conservative, it may be hoped that my general model may be found useful when more information, and more sophisticated excavation techniques, become available.

Acknowledgements : I would like to express my thanks to Dr Richard Gordon, whos contribution in the form of sound advice and textual emendation, has substantially improved the quality of this article, and to Fiona McHugh for her help.

November 2003, Lausanne

[i] Beck, R., The Mysteries of Mithras: A New Account of their Genesis, Journal of Roman Studies, 1998, 115-128 at 118.

 [ii] Daniels, C. M., The role of the Roman Army in the spread and practise of Mithraism in John Hinnells, ed. Mithraic Studies, Vol. II, 1975, 249-274.

[iii] Walters, V., The Cult of Mithras in the Roman Provinces of Gaul, Brill, Leiden, 1974. Walters remarks that only two inscriptions can be attributed to army sources. The discovery of several mithraeums in France : Gaidon-Bunuel, M. A., Les mithraea de Septeuil et de Bordeaux, Revue du Nord-Archéologie, 73 : 1991, 49-58, and Fixot, M. (éd.) Le site de Notre-Dame dAvinionet à Mandelieu, Monographie du Centre de Recherches Archéologiques 3, Paris, 1990.

 [iv] Gordon, Richard, Image and Value in the Graeco-Roman World, Studies in Mithraism and Religious Art, Variorum, Aldershot, 1996. Clauss, Manfred, The Roman Cult of Mithras, The God and his Mysteries, Edinburgh University Press, Edinburgh, translated by Richard Gordon, 2000. Turcan, Robert, Mithra et le mithriacisme, Les Belles Lettres, Paris, 2000.

[v] See Stark, Rodney, The Rise of Christianity, Princeton University Press, Princeton, 1996 ; with the discussion contained in Journal of Early Christianity 6.2 (1998), particularly K. Hopkins, Christian number and its implications, pp. 185-226.

[vi] Lofland, John, and Stark, Rodney, Becoming a World-Saver: a theory of conversion to a deviant perspective, American Sociological Review, 30, 1965, 862-875.

[vii] Stark, 16-17.

[viii] Stark, 35.

[ix] Groups which use an existing religion as a basis are more properly called sects.

[x] Stark, 193-195.

[xi] Stark, 55.

[xii] Given the different spheres of social networks that these two cults used, I do not think it necessary to bring up the idea of a competition between them.

[xiii] Beck, R., The Mysteries of Mithras: A new account of their genesis, Journal of Roman Studies, 78 (1998), 115-128 at 118.

 [xiv] Snyder, Graydon F., Ante Pacem: Archaeological Evidence of Church Life before Constantine. Mercer University Press, Macon, GA, 1985.

[xv] Beck, Mysteries,118.

[xvi] It is this lack of connection with a founding place that presents one of the enduring problems in Mithraic studies. On the other hand, the typical mithraeum closely resembles dining-rooms attached to Hellenistic temple-complexes.

[xvii] This applies only to the first groups that are archaeologically visible: the cult obviously widened its social networks within one or two generations of this period, see for example Clauss, Cult, p.23-8.

[xviii] G. Piccottini, Mithrastempel in Virunum (Klagenfurt, 1994), 44-51, has brilliantly shown that CIL III 4816 and two other fragments of a marble inscription (AE 1994: 1335) list the later entrants into the cell in the same order of seniority as AE 1994: 1334, and plausibly deduced that it marks the sole known case of the formation of a new Mithraic cell when the existing one became too large in this case, when it had apparently reached almost 100 members. It seems extremely unlikely that all those named had indeed survived, in some cases for 20 years; yet only four names are marked with the theta nigrum (indicating that they are deceased), and all of them occur in the first group, the list of the subscribers to the refurbishment of the mithraeum, whose deaths are noted in the addendum to the heading: et mortalitatis causa convener(unt).

[xix] Piccottini, 34f. = AE 1994: 1334, col. II l.10; col. III 4f. As Piccottini points out, there are good reasons why a formal album of this kind would only record the highest grade. I. Huld-Zetsche, Ein Mithräum in Mainz, Archäologie in Rheinland-Pflaz 2002 (Mainz, 2003), 75-8 at 76 fig.3, has published a fine bone ink-well from the Ballplatz Mithraeum in Mainz, rightly remarking that it shows that it was necessary in the cults day-to-day life to write documents which could not conveniently be kept on wax tablets -precisely, one may think, lists of grade acquisition and membership, as well as financial records and, no doubt, liturgical texts of the kind we possess in scraps from Sta Prisca and Dura-Europos.

[xx] Vermaseren, M. J., Mithraica I: The Mithraeum at S. Capua Vetere, Leiden 1971, 24-48.

[xxi] H.-G. Horn, Das Mainzer Mithrasgefäß, Mainzer Archäologische Zeitschrift 1 (1994), 21-66, with Beck, Ritual, passim.

[xxii]Beck, ibid. 149.

[xxiii] Beck, Ritual, 146 n. 10. 

[xxiv] Gordon, Richard, The sacred geography of a mithraeum: the example of Sette Sfere, in: Image and Value in the Graeco-Roman World, Studies in Mithraism and Religious Art, Variorum, Aldershot, 1996.

 [xxv] Carnuntum (V 457, 1682)

[xxvi] For a brief recent account, see Clauss, Mithras (n. 4 above), 131-40. Since the implication of my argument is that the grade-system was more or less universal in the cult, I will not here discuss the recent suggestions that it was effectively confined to central Italy, or was only a rather extraordinarily elaborate system of under-priests. I will just remark that the evidence of Dura-Europos is extremely inconvenient for both views, a) because the closest iconographic analogies of Dura are with the Danube area, b) because it seems clear that all initiates at Dura were in one grade or another, and in one or two cases we seem to be able actually to follow individual careers. 

[xxvii] Gordon, Roman Society, 95.

[xxviii] If, as Clauss argues, each of these grades in fact represents a priest in the cult, then there would be two priests for each grade, excepting the two highest of Pater and Heliodromus, but this does not give any indication of how many members aside from the priests could be in a cell. Given the initiation system and the close personal networks which the Mithraic cult structure seems to imply, a system of priests, i.e. elite persons having set roles and duties apart from the ordinary members, would not seem to fit the profile of a cell based organisation which allowed for growth. The post of priest would be a blocking mechanism for vertical advancement and limit the chances for ordinary members to move upwards in the cell structure. Advancement in the grades was certainly one of the incentives offered by the cult, and as mentioned before, access to more power within the organisation is one of the rewards necessary for personal fulfilment in a cult.

 [xxix] This table gives an average growth rate of 23%, which is based on the admission of 9 men to an already established cell (from table 2) assuming that each row is a 10 year period. This growth rate per decade is acceptable for a moderately successful modern cult.

[xxx] Forthcoming publication, but see review by R. Gordon in EJM online. 

[xxxii] Beck,Mysteries, 118.

[xxxiii] Stark, Rise, p. 7; idem, Modernization and Mormon Growth, in: A Sociological Analysis of Mormonism, Marie Cornwall, Tim B. Heaton, Lawrence Young eds. (University of Illinois Press, Champaign, 1994), 1-23.

[xxxiv] Beck does not attempt to quantify the founding group: the numbers proposed are my own, 100 to 1,000 men would seem to cover the description of a small founding group.

[xxxv] Stark, Rise, p. 7, table 1.1.

[xxxvi] Plutarchs description of the pirates rites is often considered insufficient proof of the introduction of the cult into the Roman world, since he says merely that the Cilician cult continues to this day (i.e. ca AD 120). But Plutarchs remark, which can hardly have been taken from his probable source, Posidonius, does reveal that an Anatolian cult of Mithras was known to him. In the absence of any other date for a founding group, Plutarch provides a convenient base line for our projection.

[xxxvii] The growth rate of 23% is taken from table 3, based on an exponential growth for a single cell of 42 men, giving 9 new members for the first ten years, 11 additional new members after 20 years, and 14 new members during the third decade. This rate is acceptable for a moderately successful cult today.

[xxxviii] Beck, in The Oxford Classical Dictionary³ p. 991.

[xxxix] Beck, Mysteries, 126-7.

[xl] Good recent examples are Künzing and Krefeld-Gellep

[xli] Quite likely this translation into stone also occurred in the case of other equipment in the mithraeum, such the cult relief panel, particularly the revolving reliefs, a technical feat in stone, but a very sensible use of material in wood. Wood survives in the archaeological record only under particular, rarely met, conditions, and given the importance of cult icons, a wooden image would more likely be destroyed than lost.

[xlii] Stark, 1996, p.39.


2006 .