The Siwa Oasis is of course most famous for its association with Alexander the Great, whose exploits have been handed down to us through the writings of his biographers Arrian and Calisthenes. In 331 BC, after defeating two Persian armies at the battles of Granicus and Issus and marching on southward into Palestine, Alexander was able to 'liberate' Egypt from Achaemenid rule. Then, for a reason not fully explained by the Hellenistic historians, the young king of Macedon suddenly had a ‘strong desire’ (Greek pothos) to undertake the hazardous journey into the Western Desert to seek an audience with the Oracle of Zeus Ammon at Siwa. This was no easy undertaking, as, on a number of occasions, the desert had already claimed the lives of other travellers who had attempted to reach this remote island of green hidden in a vast arid sea of sand covering over 680,000 square kilometres. Only two centuries earlier, a great army of 50,000 men was sent by Cambyses to "attack the Ammonians [the priests of Amun], reduce them to slavery, and burn the Oracle of Zeus" [Herodotus Book III.26]. The Persian army had tried to reach Siwa from the south, via the oasis of Kharga, but was never seen again; all 50,000, to the last man, perished on that expedition and the equipment of those unfortunate soldiers still awaits discovery, somewhere out beyond the southern oases, lying petrified beneath the drifting sand.
The Oasis of Siwa is located in the northern sector of the Western Sahara, immediately to the West of the Qattara Depresion and some 600 kilometres from the Nile Valley. The principal modern (and indeed ancient) communication route with this remote outpost is from Alexandria via Matruh (ancient Paraetorium) on the Mediterranean coast, then south across the desert. At this point the modern tarmac road gives way to an unmade track for 100 kilometres which finally reaches Siwa from the north. There are also three camel tracks leading out from Siwa: west to Girbah Oasis, north-east to Qarah Oasis and south-east to Bahriyya and from there to the other southern oases.
The Oasis of Siwa itself is a very extensive shallow basin, being approximately 82 kilometres from east to west and 28 kilometres from north to south. Within this depression are seven large salt lakes, the greatest of which are Birket Siwa and Birket Zaytun. As well as these large areas of standing saline water, which are obviously useless for irrigation, there are a number of man-made artesian wells which provide more than sufficient fresh water to irrigate vast plantations of date-palms and fruit-trees.
Rising out of the greenery are a number of rocky outcrops on which, since time immemorial, the local population groups have built their fortified settlements, affording them protection against the more dangerous desert fauna and, in particular, nomad/Beduin raiders.
Siwa Oasis with Siwa Hill in the background
The modern inhabitants of Siwa are mainly of Medieval Berber descent, having arrived in the region in the early 13th century in a group one hundred or so strong. It is only in relatively recent times that they have relinquished their unusual Berber dialect for modern Arabic. The forty men and their families settled around a rocky cone, building a fortified stronghold which they called 'Shali' (Berber for 'town'). This was later to become 'Siwa', after the name of the particular Berber tribe who had founded the settlement, and now forms the heart of the modern village of that name. The pharaonic name for the oasis is not certain but what limited evidence there is suggests that it may have been called 'Thay' or 'Tha'.
The medieval village on Siwa Hill
The more ancient inhabitants of the oasis were settled on and around the rocky crag of Aghurmi, some four kilometres east of Siwa town, and they continued to reside within the walls of this mudbrick fortress until 1926 when a heavy thunderstorm (lasting three days) destroyed a large number of houses, killing their inhabitants. The population then departed from the site, which had been occupied since pharaonic times, and built a new settlement nearby. It was that exodus from the crag which enabled the archaeologists to begin the work of clearing the medieval occupation in order to gain access to the more ancient remains which had been identified by James Hamilton in 1853 but which were otherwise only known through the stories of Alexander's visit to the Oracle.
The Ancient Pharaonic Sites
There are a number of ancient sites within the Siwa depression and around its periphery, including Aghurmi, Umm Ubayda, Ayn el-Gubba, and Gebel el-Mawta. Of these the settlement atop the Aghurmi crag is certainly the most interesting because it was here that the Temple of the Oracle of Amun was rediscovered in the last century and where most of the subsequent excavation and survey work has been concentrated in recent years. Before turning to this site, which will form the basis of this discussion, we should perhaps first take a brief look at the other locations containing pharaonic remains.
The site of Umm Ubayda includes what was once an impressive temple built by the 'the Great Chief of the Desert', Wenamun, for the pharaoh Nectanebo II (360-343 BC) of the 30th Dynasty and dedicated to Amun. The temple lies within the palm groves just half a kilometre to the south of Aghurmi.
Unfortunately, and not untypically, this fine monument was blown up by a local official, one Mahmud Azmi, in 1897 so that the blocks could be used to construct a new staircase for the local police station and, needless to say, one also for his own house. What remains today is the stump of a wall from the sanctuary which never the less is richly decorated with scenes of Wenamun, Libyan feather in hair, kneeling before Amun and with a procession of deities behind him. The two temples of Aghurmi and Umm Ubayda are connected by a ceremonial way and probably functioned under a single administration, given the isolation of their setting.
Alexander's biographers mention a second temple of Amun, other than the Temple of the Oracle, and describe its setting amidst the palm groves of the oasis; this is no doubt the temple at Umm Ubayda.
Ayn el-Gubba is the famous 'Spring' or 'Fountain of the Sun' to which Herodotus accords miraculous powers: he tells us that whilst during the hot hours of daylight its waters remain cool and refreshing, at night, however, the spring "boils furiously" [Herodotus Book IV.182]. In fact, the bubbling effect is caused by natural gasses rising from rocks beneath the surface – the temperature, of course, remains much the same throughout the 24-hour day.
The limestone cliffs of Gebel el-Mawta, a promontory rising out of the palm groves in the northern sector of the fertile basin some five kilometres north of Aghurmi, formed the principal burial ground of the ancient Siwan settlement. Indeed, Gebel el-Mawta means 'The Hill of the Dead' in Arabic and is also called Qaret el-Missabbarin – 'The Ridge of the Mummified'. There are a great number of tombs in the area, many of which are now completely buried in sand. A number of others will have remained undiscovered, since a systematic excavation of the site remains to be undertaken. The important published tombs are those of Niperpathoth, Mesuisis, Siamun and the so-called 'tomb of the crocodile'.
The most interesting is that of Siamun which has fine colourful scenes of the deceased and his family painted on plastered walls. There is little doubt that, in spite of his Egyptian name, Siamun was not of Egyptian stock as he is shown with pale complexion, beard and thick black curly hair. His wife is darker skinned and bears the good Egyptian name Neferheret – 'beautiful of face'; she most probably, therefore, was a native Egyptian. Siamun's father, with the name Periytu, is, however, undoubtedly of foreign extraction. As if to add further confusion, the son of the tomb owner is depicted in Greek dress whilst Siamun himself wears the clothes of an Egyptian nobleman. Thus the tomb of Siamun is strong evidence for the international character of Siwa during the Hellenistic Period and appears to demonstrate that intermarriage was practiced between the Greek world and Egypt.
We now turn to the most important of the Siwan sites, that of the Temple of the Oracle itself. The pharaonic remains are located on the relatively flat-topped rock of Aghurmi which forms a natural acropolis rising to some thirty metres above the palm groves. It appears that, not long after the temple and ancillary buildings fell into disuse, the rock was occupied by the local inhabitants who moved from the vulnerable low-lying areas into the fortified citadel of the Siwan chieftains – the local governors of the pharaohs. As a result, when the temple was discovered in modern times it was sheathed in mudbrick houses with very little of the ancient structure visible.
The major task of clearing the temple of its Medieval overcoat was undertaken by Ahmed Fakhry in 1970 but before that three brief surveys had been completed, first in 1932, by Ricke, Steindorff and Aubin, and then in 1938 and 1941 by Fakhry himself. More recently, during three seasons of activity at Siwa, a much fuller investigation of the temple and its surroundings was conducted by the German Archaeological Institute in Cairo under the direction of Klaus Kuhlmann. This work has revealed some very interesting details about the workings of the Oracle and its temple which had not been hitherto noted and these findings will be dealt with at the end of this section on the temple at Aghurmi.
During the two days that Ricke and his colleagues spent at Aghurmi they attempted to establish the building date of the Temple of the Oracle and indeed Steindorff, whose assignment was the investigation of the 'Holy of Holies' or 'cella', did locate a much damaged cartouche which he read as that of Achoris (393-381 BC), third king of the 29th Dynasty. Thus the earliest known building activity at the site – the sanctuary itself – could be taken back only as far as the post-Persian era. This was obviously very disappointing, as it cast doubt on the veracity of Herodotus' story about Cambyses' attempted assault on the Oracle. Then, in 1938 Fakhry cleaned down the inner walls of the cella and was able to show that the builder of the temple was in fact Amasis (570-526 BC) the last ruler of the 26th Dynasty, just prior to Cambyses' invasion. Thus Herodotus was vindicated by modern archaeology for now there was proof of the existence of the Oracle in late-Saite times.
The Temple of the Oracle
Let us now take a more detailed look at the Temple of the Oracle to learn something of the way it was built and how it may have functioned, at the same time recalling the Alexander-narrative so as to determine the accuracy of the Hellenistic descriptions.
At first glance it appears that the temple is of the basic Late Period/Pre-Ptolemaic Egyptian design, consisting of a courtyard and gateway with an interrupted lintel leading to an open portico/hall, a second inner hall and finally the central sanctuary with chambers on either side. However, when one looks further, there are a number of very untypical aspects in terms of both the temple's location and its design.
The Temple of the Oracle of Amun on Aghurmi Hill
Firstly, and perhaps most obviously, the temple is located on top of a rocky crag. Throughout the whole corpus of temple or shrine building activity in ancient Egypt this is almost, but not quite, an exceptional occurrence. Only four other examples come to mind, the last two of which are even then somewhat doubtful. The obvious and most clear cut example is the shrine of Thoth built by Saankhkare Mentuhotep (2010-1998) of the 11th Dynasty which is located atop the highest peak in the hills of Western Thebes. Secondly, there is the Ptolemaic temple of Petesuchos at Karanis in the Faiyum – not only later than the Siwan temples but perhaps also the closest in character to the Temple of the Oracle. Then we may tentatively add the 'high place' at el-Hiba, which, although probably a sacred altar of some kind, now has insufficient remains to indicate if there was ever a mudbrick building at the site. Finally there is the Predynastic temple/shrine of Horus at Hierakonpolis which appears to have been erected on top of a large platform. All the other temples of dynastic Egypt (apart from the rock-cut speoses), as far as I am aware, were built on low-lying ground. In this respect therefore the Aghurmi temple is most unusual.
Furthermore the temple was reached by a grand stairway leading up from the base of the rock where the sacred well was located (mentioned by the Greek historians). This is only paralleled at Karanis but, even then, on a much smaller scale.
At the head of the stairway there was a narrow terrace in front of the entrance to the court. This court may have been that described in Alexander's story where the barque of the Oracle was paraded around the king and his escort before Alexander went alone into the shrine for his private audience. Above the low wall, on either side of the main entrance gateway, a second facade had been erected in Greek times, perhaps at the behest of Alexander himself. Attached to this second wall, and standing on the cavetto-cornice of the earlier 26th-Dynasty facade, were two fluted half-columns. The facade was otherwise completely plain and without inscription.
Now we come to perhaps the most intriguing features of the temple, the ones which give it its special character as the seat of one of the most famous oracles of the ancient world.
The cella, or shrine of the Oracle, possessed a number of special features which enabled the priests to put on dramatic performances for those wishing to seek the wisdom of the Oracle. Firstly, the cella had a second storey, which had no visible entrance, from where a priest could secrete himself and deliver the words of the ram-headed statue of Amun. Thus the source of the voice of the Oracle remained heard but unseen by those who had been granted a private audience – as in Alexander's case. The Oracle's theatrical performance could be further enhanced by other priests shaking sistra, making noises, or playing the roles of the other gods. This was done by means of another secret corridor located behind and to the right of the statue. Thus the voices of the priests could be made to resonate through the walls of the shrine echoing in the darkness all around the statue of Amun, bedecked in jewels – the ancient equivalent of a Son-et-Lumiere performance at Karnak perhaps? This corridor was only accessible via an ingenious tunnel, the entrance to which was only recently found at the base of a cliff immediately to the rear of the temple on the northern edge of the acropolis. By this elaborate system of hidden rooms and corridors the priests were able to impress even the least impressionable of visitors, and it is not really surprising that the Oracle of Zeus-Ammon was so renowned during the Greco-Roman era when exotic cults and mystic religions became so popular.
What we do learn from this, however, again leads us to think how un-Egyptian the temple and its Oracle were by the time Alexander made his visit. It is reasonably clear, from the archaeological remains of the pre-Ptolemaic temples of Egypt, that the Egyptian clergy did not go in for elaborate deceptions – in particular ones which involved the performance of the god's spoken word. We know this because no hidden chambers have been found earlier than that at Siwa. It is only in the Greco-Roman temples such as Dendera, Edfu and Karanis that such devices are introduced to deliver the oracular messages. We further observe that the normal method of oracular delivery in true Egyptian style was the silent nodding of the sacred barque of the oracle as it was brought before the questioner. This ceremony was usually performed not inside the holy shrine but outside in the public gaze. Indeed, as we have noted, this is precisely what happened when Alexander and his men were first introduced to the Oracle in the forecourt of the temple, before the king was given his private audience in the inner sanctum where he learnt of his divine parentage.
It thus appears that the oracular voice was a Greek invention and that the building of the Aghurmi temple, late in the 26th Dynasty, may have been a partially Greek affair, undertaken by an expeditionary force of Greek mercenaries sent by Amasis to garrison the Oasis. The involvement of the local native ruler of the Siwa Oasis, a Libyan named Sutekhirdes 'Chief of the Desert Dwellers', is quite evident as it is his figure which occupies one of the inner walls of the shrine, offering to the gods, whilst the king is portrayed on the opposite wall in a similar fashion.
Although Alexander's experiences at Siwa are the only written records that we can draw upon to learn something of the history and nature of the place, modern excavations and surveys have revealed further information which not only confirms the accounts of Callisthenes and Arrian but also appears to show that Siwa only came under Egyptian control in the Late Period.
The earliest pharaoh recorded at the Oasis is Amasis and none of the other buildings, including the outlying temples of Bilad el-Rum, Qasr el-Ghashsham, Abu Sharuf and el-Zaytun (not discussed in this essay), are earlier than 360 BC. Indeed all are of Greek or Roman date.
The local chiefs or governors whose images appear on the Aghurmi and Umm Ubayda temples, along with Amasis and Nectanebo, do appear to be of Libyan stock and one must presume that contact was made between Egypt and Siwa certainly by the 22nd Dynasty. Siwa may have in fact been one of the primary staging points in the Libyan migrations of the early Third Intermediate Period.
With Old Kingdom mastabas having been found at Dakhla, we know that the Southern Oases were under Egyptian influence for most of the period of dynastic rule, but no such evidence exists for Siwa. Acknowledging that the following assumption is based on negative evidence, nevertheless I have to conclude that the Oasis of Siwa was essentially a foreign land and not a part of Egypt until the area was brought under pharaonic control in the late-26th Dynasty of the Saites. Subsequent archaeological investigation may invalidate this statement, however, for the moment, this conclusion, having been founded on what the site has revealed to date, appears to be justified.
C. D. Belgrave: Siwa: The Oasis of Jupiter Ammon (London, 1923).
A. Fakhry: Siwa Oasis: Its History and Antiquities (Cairo, 1944).
--- 'Recent Excavations at the Temple of the Oracle at Siwa Oasis' in Agyptischen Bauforschung und Altertumskunde, Vol. 12 (Wiesbaden, 1971), pp. 17-33.
--- The Oases of Egypt, Vol. 1 (Siwa Oasis), (Cairo,1973).
K. P. Kuhlmann: Das Ammoneion: Archaologie, Geschichte und Kultrpraxis des Orakels von Siwa (Mainz an Rhein, 1988).
W. J. Murnane: The Penguin Guide to Ancient Egypt (London, 1983).
G. Steindorff: Durch die Libysche Wuste zur Amonsoase (Leipzig, 1904).