Religions in Berenike
by: Robert E. Hughes

One of the key questions regarding the ancient port town of Berenike has been the religious character of the population of the site through the established eight centuries of occupation. Berenike was a conduit for goods coming in from sub-Saharan Africa as well as the southern coast of Arabia and India to be passed through to the Nile Valley, and from there through Alexandria to a variety of points throughout the Mediterranean and vice versa. Due to this nature and the alluring possibility of economic gain, a cosmopolitan mix of individuals can be expected, ranging from Egyptians to Greeks to Romans to Indians to local nomadic peoples, anyone interested in making a profit. Documents uncovered during excavation thus far indicate the presence at various times of individuals of Latin, Greek, Egyptian, Hebrew, Aramaic, Palmyrene, Tamil Brahmi, north or south Arabian and of Ethiopian descent. It can be hoped that examination of the remains of religious cult practices unearthed at Berenike will shed light upon exactly which of these peoples actually lived there, and who may have been transitory or not present at all. The information found in the last seven seasons of excavation have yielded some direct evidence as well as some oblique references which pose more questions than are answered. Regardless, the information is rich and extremely valuable in terms of understanding not only Berenike but the climate of Graeco-Roman Egypt as a whole.

The first area in which evidence for cultic activity was found is trench BE95/96-6, the finds of which prompted the opening of trench BE97/98-16. It has subsequently come to be referred to as the 'West Shrine' to distinguish it from another cultic structure found in later seasons referred to as the 'North Shrine' which will be discussed later in this text. The finds within the West Shrine have been curious indeed. Excavation in the two adjoining trenches yielded a fairly complete stratigraphic profile of the site from the Ptolemaic period to the late Roman era in the fifth to sixth centuries AD. The earliest phase reached was the late Ptolemaic pottery dump into which the early Roman phase had cut. However, the depth of the trench prohibited safe excavation beyond the late Ptolemaic phase. Boring results suggested that Ptolemaic structures exist below the late Ptolemaic pottery dump. No substantial architecture remains of the early Roman construction in this trench and the third phase dates to the third or fourth century AD and has been called a pagan shrine. This phase was found to exhibit four major architectural overhauls.

The first sub-phase seems to have been a cult to the Palmyrene god Hierobal or Yaribol combined with the Roman imperial cult. Finds consisted of a monumental stone block with two footprint shaped cuttings in the northern end of the block. This suggests that a nearly life-size statue was once installed there. The reason that Hierobol is suggested is from the discovery in an upper layer of a reused block inscribed to this particular god. The inscription had fallen out of the baulk of trench BE95/96-6 during the break between seasons and is the primary reason that the decision to open adjacent trench BE97/98-16 was made. The inscription is bilingual written in both Greek and Palmyrene and provides the name of the sculptor, Berechei; names the official in charge of Berenike, Amelius Celer, Prefect of Mount Berenike and of the Ala Heracliana; and attributes the inscription to the god Hierobal/Yaribol. Presumably it would have accompanied the statue which was found in the same trench earlier. The probably date of the inscription is between 180/185 and 212 AD. The Ala Heracliana was not transferred to Egypt until after 180 AD and the fact that Berichei does not have an 'Aurelius name' suggests that it was written before the Constitutio Antoniniana of 212 AD which made all free inhabitants of the Roman Empire citizens. Those individuals without the full complement of names usually took the name of the emperor out of honour and respect.

The Ala Heracliana was probably stationed at Palmyra before being transferred and thus would consist primarily if not totally of Palmyrene soldiers which would explain the dedication to a Palmyrene god and for the good fortune of Palmyra. The pottery associated with this stratigraphic level corresponds with the dating of the inscription and allows this to be assigned to the late Antonine and Severan periods with a fair degree of confidence, that is at least the second half of the second century AD and probably later. In addition to this inscription, another was found in trench BE95/96-6 in a later layer which was dedicated by another Palmyrene, Marcus Aurelius Mocimos, to the Roman imperial cult of Caracalla and his mother Julia Domna. This insciption was found on two blocks which were found tossed about the trench resulting from a robbery which occurred between the 1995 and 1996 seasons. But it is with a fair degree of confidence that they were part of the podium upon which a gypsum base and bronze female statue rested. The inscription is clearly dated 8 September 215 AD, during the twenty-fourth year of the reign of Caracalla. Here we see that Mocimos has taken the nomen and praenomen of the emperor as his right and privilege guaranteed by the Constitutio Anoniniana of  212 AD. The inscription itself does not mention the statue but the measurements fit and the reconstruction of the inscription into the statue podium seems accurate.

That the podium and its statue and insciption post-date the inscription and statue base for Yaribol/Hierobol despite the Palmyrene connection between them is evident by the discovery of a heavily looted but still substantial flagstone flooring at the base of the statue podium which lies above the stratigraphy of the earlier insciption and associated statue base of Berechei. The statue itself is fragmentary but about 60 % of it was recovered as pieces scattered throughout the trench. The statue is preserved complete from just above the knee down. She wears a peplos type garment down to and covering the feet and rests its weight on its right leg with its left leg bent at the knee. This stance is typical for a Graeco-Roman statue. No head was found but the left arm and clenched left hand were recovered. The hand appears to be holding either a snake or possibly a cornucopia. Either item is suggestive of a deity, the identity of whom is still in question. The possibilities for the identification are numerous. If the item in the hand is a snake, then this could be Hygeieia, daughter of the Graeco-Roman healing deity Asklepios/Aesculapius. She is frequently depicted either holding or associated with a snake. If it is a cornucopia, then Isis is a likely possiblity. She is rarely depicted holding a snake but often with a cornucopia and was incredibly popular in Graeco-Roman Egypt. Isis was also syncretized with many other classical deities as well as female members of the Ptolemaic royal house including Arsinoe Philadelphus, sister and wife of Ptolemy II Philadelphus, the founder of Berenike. Isis also had connections with sailing and Greek hymns frequently referred to her as a goddess of the sea and navigation. As such the syncretism Isis/Tyche is possible, several representations of which are known from Alexandria. Because of the fluidity of Egyptian syncretism, the combination Isis and Hygieia is also not out of the question. The cornucopia could also suggest a syncretism of Isis and Demeter, also attested from Graeco-Roman Egypt. However, because of the imperial cult inscription, this could well be an image of Julia Domna or possibly an image of the Ptolemaic ruler cult. According to the excavators the style of the statue appears much earlier than the fourth-fifth century AD date ascribed to it by the associated ceramics. In addition, the statue does not appear to be on its original base but may have been reused here from elsewhere. Both these facts would support the identity as a Ptolemaic queen or Ptolemaic queen as Isis. Isis is the common denominator of the majority of the options available and it is hoped that the other cultic items found in association would make the determination clearer.

Concurrent with the flagstone flooring of the shrine in trench BE95/96-6 a gypsum altar and sphinx and an ashlar podium were found in trench BE97/98-16. All three were covered in an ash layer which most likely represented debris from earlier offerings burnt on the podium. Two decayed wooden bowls found in association with the sphinx and podium suggest that the bowls were involved in the offerings but this was made certain by a second gypsum altar which emerged from the baulk in association with these materials. This second altar preserved the remains of the bottom of a small wooden bowl set into its upper surface on a bed of white gypsum paste.  Sectioning of the bowl on top of the altar revealed evidence for how the wooden bowls were used in cultic activity. The following is one excavator’s hypothesis. First, a wooden bowl is set into position on the altar on a bed of white gypsum paste which had reversible adhesive properties. Then, a lining of sand was placed into the bowl. This served as a protective layer to prevent the bowl from burning through. Charcoal from a brazier was placed inside the bowl. A number of bowls containing sand and charcoal as well as several broken pottery vessels full of charcoal were found in association with the wooden bowls. These may have been used to measure out the required amount for an individual offering. All these items appear to have been originally stored on wooden shelves upon the walls of the room which dumped their contents to the floor during the decay after the structure was abandoned. This would also help explain the remains of over 100 wooden bowls found along the walls of the structure within the windblown accumulated debris from the abandonment phase. It appears that they were stacked and stored on shelves upon the walls in preparation for worshipper use. The sheer number of bowls suggests that participation in cultic activity was not restricted to a priestly group but possibly engaged in by the general public. It should be noted that this idea of an individual participation in cultic activity and worship is strengthened by the find of similar bowl remains in trenches BE95/96/97/98/99/00-10 and BE99/00-23 north of the Serapis temple in the center of town.

At some point this shrine seems to have fallen out of use and the Palmyrene inscription covered in accumulated debris. There may have been a cultic change but the female statue seems to have been included in this change. This late fourth to early fifth century AD use of the area yielded another collection of cultic materials. These include a gypsum lion’s head, a possible stela base with lead lining which was stolen during the robbery already mentioned, a rectangular basin carved from a column fragment, several gypsum slabs with unknown function, two tentative incense burners without evidence of use, and a gypsum offering table in the shape of a temple pool or sacred lake with stairs on two sides of its interior. Offering tables in general are an Egyptian rather than a Hellenistic phenomenon. This type is documented from the Ptolemaic period onwards, especially in Upper Egypt and Nubia. Of all these finds, this one promises to be the most valuable in determining the cultic function of the shrine because more of these basins were found in the North Shrine in later seasons.

The North Shrine was encountered in trench BE98-23, which is located just west of a structure known as 'Building F'. Several trenches during several seasons were undertaken in Building F because its size and lay-out with clear evidence of a temenos wall suggested a temple. But nothing was found which would indicate any religious function of this structure with the possible exception of some painted ostrich egg shell fragments found in trench BE95/96-7. The decision to open BE98-23 was motivated in part by a desire to help clarify the nature of the activities of this area of the site. The structure revealed two periods of use. The later phase revealed a circular gypsum altar which was surrounded by ashes. It had a slightly concave top with burned material inside probably remnants of a burned offering, possibly similar to the charcoal dedication remains found in trench BE97/98-16. However, no wooden bowls or their remains were found in this trench. There was also a small offering table consisting of a rectangular stone block which also had a circular burned patch on it. Near the circular altar a complete ostrich egg painted with designs in red was found. And to the immediate north of the altar, three temple pool offering basins were revealed and a fourth one to the east of the altar. All the basins except one appear quite worn which suggests they had been in use for quite some time. The trench also contained a number of broken terra-cotta veessels including some amphorae and wooden objects such as a teak box and the remains of wooden structures possibly shelving units made of teak, local acacia and imported conifer. The teak structures are probably secondary usage for which there are numerous parallels throughout the site.

It is the presence of the temple pool basins which are most intriguing. Three of the four have a flight of stairs carved into each of the four internal sides. These three have a double incised lines around the rim of the basin which extend onto a kind of spout. They may have served some sort of controlled drainage to the spout. These same three exhibit wear marks along the back, back left and right and right respectively (back referring to the side opposite the  spout which was presumably the front). The basin which is not incised nor exhibits any wear patterns is also devoid of any carved staircases. This could suggest that it was unfinished at the time of deposition or abandonment, although its placement in the middle of the line of three would not indicate a work-in-progress. None of the basins are the same dimensions, the smallest measuring 150x140x50 mm and the largest measuring 250x220x100 mm, and thus feasibly portable sizes. Parallels for this type of offering or libation basins are extant from Koptos, Denderah, Shabul with Meroitic inscriptions, and another from Qasr Ibrim. Unfortunately the exact purpose and usage of this type of offering basin is not fully understood.

In order to better understand the nature of the cultic activity in this trench, excavation continued down to the first level of occupation and the trench was extended to the east with adjoining trench BE99/00-32. The structure revealed was a rectangular building oriented east-west with a single door jamb in the east wall. The circular altar discussed previously is part of the original equipment of the structure and its use spanned both phases. It is centrally located at the rear, or western end, of the building. In addition, the remains of wooden bowls for burned offerings similar to those found in BE97/98-6 were found to the left of the doorway inside the structure. From these finds the building has been deemed a temple with confidence. This is the extent of the evidence for cultic practice within this temple and the deity to which it could be ascribed is as yet undetermined. One excavator has proposed the hypothesis that this temple could be a Mithraeum, a place of worship of the god Mithras. The popularity of this cult of eastern origin, especially among Roman soldiers, was only rivaled only by the cults of Isis and Christianity. The only support for this theory, however, is some architectural features of the temple. In general the features of a Mithraeum are that it is a rectangular structure with benches on the two long sides if not three sides excluding only the point of entry. They are always vaulted structures and generally are located underground. The far end of the structure has a niche in which the cult image of the god is located, invariably a stone representation of a youthful male involved in the act of slaying a bull. An altar or altars might be present as well as a basin before the cult image and a drain channel may be present but is not a regular feature of Mithraea. The temple at Berenike called the North Shrine exhibits a couple of these features but some key ones are notably absent.

The strongest parallel with known Mithraea is the fact that the central area of the room is set apart from the exterior walls by a barrier of fossilized coral heads running parallel with the long walls (north and south) and a beam of teak wood set parallel to the west all opposite the entry wall. It should be noted that coral heads were the preferred medium of construction throughout the site. The reuse of teak from Indian ships has already been discussed. These barriers create a meter wide band around the interior of the structure suitable for the benches expected in a Mithraeum. However, no benches were found. But the area was devoid of cultic ash and debris found in the central space and there is evidence that matting may have covered the perimeter area. The existence of the centrally located altar has been discussed. However, there is no basin or drainage channel. There is also no niche at the rear wall of the building in which a cult stature would have been placed, nor any evidence of a cult image. A bottom corner piece of sculpted fragment was found which exhibits part of one male figure and perhaps part of the leg of another which could be the leg of Mithras and one of the tunic wearing twins who normally flank the divine image. Although tantalizing evidence, the fragment was found as part of the stone barrier which partitioned the room and not in a context which would allow its interpretation as a cult image. Lastly, the structure is not underground nor even partially submerged and there is questionable evidence concerning a vaulted roof. One excavator claims there is an indication of curvature of the long walls indicative of vaulting. However, most Mithraea exhibit vaulting roughly 2 meters from the floor and the extant walls at Berenike do not reach this height in this structure. Buttressing would also be expected in order to support the outward thrust of such a vault and there is no evidence of any such measures at this time.

Despite the lack of direct evidence for the presence of Mithras at the site, and in Egypt for that matter, Mithras does have ideological and iconographic associations with cults present in Egypt. Mithras is associated with Deus Aeturnus who is attested at Oxyrhyncus near the entry to the Fayoum. This deity in turn is iconographically linked with Sol Invictus who is attested in a victory relief of the Meriotic King Sherkarer. Berenike undoubtedly had contacts with Nubia and the nomads who populated the desert between Berenike and the kingdoms to the south, but it is a weak link indeed to propose the presence of Mithras through this link. A more likely explanation could be found in understanding a more localized form of Mithras or Sol Invictus emanating from the connections with Nubia. The Nubian god Mandulis, for whom the temple of Kalabsha was contructed by the the Ptolemaic kings and Roman emperors attests to the strength of the Nubian cults in Egypt at this time. It is equally plausible that this temple could be for a Nubian deity as for Mithras. After all, the existence of a shrine to a Palmyrene god was not one that was expected to be found. The enigmatic offering basins would support the attribution to an Upper Egyptian or Nubian cult if they can be ascribed to the same deity. As with trenches 6 and 16, they belong to a later phase of use of the structure and it is uncertain if the deity worshipped therein remained the same. Moreover, a thorough study of the use of these basins needs to be undertaken if an answer is to be attained.

In sum, the evidence for religious cult practice at Berenike to date has answered some questions and provoked many more. In addition to the state cult of Serapis whose stately temple dominated the center of the site, there is evidence at different periods for the worship of a Palmyrene deity; the worship of a female deity of unknown identity although the most likely candidate is Isis in one of her many syncretized forms including the Ptolemaic ruler cult or perhaps the Roman imperial cult; and the worship of an unknown deity whether Mithras, traditional or localized in form, Sol Invictus or a Nubian cult. We also have two known areas of cult activity within the site other than the known temple of Serapis at the center. Most importantly, the evidence provided by the excavations at Berenike have provided direct evidence of individual participation in cult activity and exactly the nature of that cult activity through the burning of incense or burnt offerings by means of the copious wooden bowl remains, ‘offering-kits’, by which persons other than the priestly class gained access to a deity. Such evidence is unparalleled and is an invaluable contribution to the understanding of religious cult practice at Berenike and Graeco-Roman Egypt as a whole.

Back to Introduction
Back to Contents